2006-FuranicCmpdsAllAspects

advertisement
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/310319080
PRACTICAL EXPERIENCE GAINED FROM FURANIC COMPOUND ANALYSIS
Conference Paper · April 2006
CITATIONS
READS
12
297
1 author:
Lance Richard Lewand
Doble Engineering Company
20 PUBLICATIONS 152 CITATIONS SEE PROFILE
All content following this page was uploaded by Lance Richard Lewand on 15 November 2016.
The user has requested enhancement of the downloaded file.
PRACTICAL EXPERIENCE GAINED FROM FURANIC COMPOUND ANALYSIS
by
Lance R. Lewand
Doble Engineering Company
INTRODUCTION
When cellulosic materials are heated above 100°C due to any number of reasons, they begin to generate
characteristic degradation byproducts, some of which are oil soluble. These can be sampled easily and
used as indicators of aging. These types of tests are termed “indirect tests” as they are not direct
measurements on the paper. There are two indirect tests: dissolved gases-in-oil; and furanic compounds in
oil. This paper focuses on one of these indirect tests, that being for furanic compounds.
Furanic compound analysis has been used for close to twenty years to aid in the diagnosis of cellulosic
insulation in dielectric liquid filled electrical equipment. Yet, it is not as well understood as dissolved gasin-oil analysis (DGA), nor are the factors that influence the generation, degradation and accumulation of
furanic compounds.
This paper explores factors such as, degassing, fluid change out, reclaiming and
reconditioning the oil and other maintenance activities that influence the concentration of furanic
compounds in dielectric liquid. In addition, it explores furanic compound accumulation patterns in
different types of transformer systems, namely core and shell type design; thermally upgraded, regular
Kraft, Nomex®, and mixed paper/ Nomex® insulation; and various means for oil expansion and
preservation.
Mineral oil, paper and other cellulosic materials such as pressboard and wood are the primary components
of the transformer insulation system. These insulating materials have been selected because of the
following qualities: 1) abundance resulting in low cost, and 2) reasonable longevity under normal operating
conditions. However, these materials will degrade over time and therefore have a finite life. As a result,
the life of the solid insulating materials is directly related to the life of the transformer. Because of this
great emphasis has been placed on estimating residual insulation life in transformers.
DEGRADATION OF PAPER INSULATION
Thermal stress and the concentration of water and oxygen influence the degradation rate of insulating
materials. The most critical component of the insulation system is the paper intimately wound around the
copper or aluminum conductors in the windings and therefore not easily replaced. Good quality mineral
insulating oil is expected to last 30 or more years before forming excessive amounts of acids and sludge.
Although important, is not as critical as the cellulosic insulation because it is easily reconditioned to
remove water and particles, reclaimed to remove degradation products or replaced. Therefore, the
longevity of cellulosic materials, becomes the limiting factor in the operation of transformers [1].
As paper and other cellulosic materials deteriorate, byproducts such as carbon oxide (carbon monoxide and
carbon dioxide) gases and furanic compounds are formed. These compounds can serve as indicators to act
assess the aging process. The degradation of cellulosic materials can also be measured directly by the
degree of polymerization (DP) test. Before discussing the relative merits of each of the tests, it is useful to
review aging factors and the chemistry of cellulosic materials, deterioration processes, and byproduct
formation.
AGING FACTORS OF CELLULOSE INSULATION:
The effects of temperature, water, and oxygen are significant factors in the aging of the paper insulation
(cellulose) and oil. Aging processes have been explored extensively through accelerated aging tests and
field experience. A general discussion of the effects of aging factors on paper insulation is provided.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
1
Effects of Temperature:
In general it can be stated that the primary cause of deterioration of paper is from thermal instability.
Thermally-upgraded paper is less sensitive than Kraft when exposed to excessive temperatures. As Mr. W.
J. McNutt [2] discussed in his paper, the aging of the paper insulation follows Arrhenius type-kinetics.
Although he expressed the effects of temperature on aging with an equation, a rough index is that for every
6 to 8°C rise in temperature, the life of the paper insulation is halved. Of course, the temperature at which
you begin is important. For instance, using Mr. McNutt's equation, if the operating temperature of the
insulation is 40°C, the life of that insulation is approximated at 110,000 years. However, if this same
insulation is exposed to a temperature of 140°C the estimated life is now only about one year [2].
Effects of Water:
The effect of water on the aging of paper is significant and deleterious. The rate of paper degradation is
directly proportional to the water content. For example, decreasing the water content of the paper from
1.0% to 0.5% doubles the life of that paper. Thermally-upgraded paper insulation is less sensitive to the
effects of water than Kraft paper.
Effects of Oxygen:
Paper aging is influenced by the presence of oxygen although not to the same degree as oil. Thermallyupgraded paper is even less sensitive to the effects of oxygen than Kraft paper. It has been suggested that
the effects of high oxygen compared to low oxygen environments on the aging of Kraft paper is of the
order of 2.5 to 1 [2], which seems reasonable (based on our own experience in examining paper samples
from service-aged transformers).
The expected life of dry (≤0.5% water) regular Kraft paper in an high oxygen environment (such as very
old free breathing designs) is about 4 years performing at a temperature of 100°C (the expected hotspot
temperature of a transformer at nameplate rating of 55°C rise). In contrast, the expected life of dry,
thermally-upgraded Kraft (TU) paper in a low oxygen environment (sealed conservator or nitrogen
blanketed) performing at a temperature of 110°C (the expected hotspot temperature of a transformer at
nameplate rating of 65°C rise) is about 18 years [3]. It becomes apparent then, that the types of paper
insulation and preservation system severely impact the expected life of the cellulosic insulation. Even so, if
paper insulation is maintained in a dry state it will retain good electrical properties even as it becomes quite
brittle. Through the course of winding vibration and movement however, particularly during through
faults, mechanically weakened paper can break, reducing insulating capability or allowing conductor
movement and reduced clearances. It is then that dielectric failure is more likely to occur.
CHEMISTRY OF CELLULOSE
The pulping process for electrical Kraft paper converts the wood chips to cellulose by removing the
majority of lignin (95-98.5%) and other impurities [4]. The main pulping process used today and the one
that is used to produce electrical grade Kraft papers is the sulfate process which is also called alkaline
pulping. Sodium hydroxide and sodium sulfide are used in what is termed the cooking process. The
cooking process under conditions of heat, pressure and chemicals (pulping liquors) removes the lignin and
impurities from the wood chips in order that only cellulose remains. The pulping liquor is removed and
recycled for use again and the remaining cellulose pulp is washed several times to remove as much of the
pulping liquor as possible from the cellulose pulp. The Kraft process is slightly different in that the same
chemicals are used but the pulp is intentionally undercooked and results in the darker color of the paper as
well as exceptional mechanical strength.
Cellulose (Figure 1), known to be the major constituent of paper and pressboard, is a long straight chain
polymer (polysaccharide) of glucose molecules. Glucose is a sugar that has 6 carbons (C), and is typically
in the more stable ring structure called a pyranose. The glucose rings are linked at ring positions 1 and 4 by
an oxygen atom in what is referred to as a glycosidic linkage or bond.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
2
FIGURE 1
The long-chain cellulose molecules interact with each other due to hydrogen bonding. This occurs because
the outer electrons are held more closely by oxygen (O) than the hydrogen atoms (H) covalently bonded in
the hydroxyl groups (OH), thereby imparting a small polarity. Hydrogen atoms are then attracted to the
oxygen atoms on adjacent molecules, resulting in hydrogen bonding (Figure 2). This provides cohesiveness
to the cellulose molecules, resulting in strands, mats, etc.
H
H
O
O
C
C
Dashed Lines Represent Hydrogen Bonds
FIGURE 2
Much of the mechanical strength of paper and pressboard comes from the long-chain cellulose polymer.
As the cellulose ages, the polymers are cleaved and become shorter, resulting in reduced mechanical
strength. The primary forms of degradation are thermal, hydrolytic, and oxidative, and as seen in Figure 3
[5] that results in a reduction in DP. In the case of each of these mechanisms, free glucose is generated and
the ring structure tends to be opened to form chains. Although temperature is likely to be the most
important factor, oxygen and water have been clearly shown to have a significant effect on the degradation
of Kraft paper.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
3
HYDROLYTIC, OXIDATIVE AND THERMAL DEGRADATION REACTIONS
FIGURE 3
The degradation of cellulose molecules results in the formation of gases, primarily carbon monoxide and
carbon dioxide, glucose which in turn forms furanic compounds, and other byproducts. Furanic
compounds are 5-membered ring structures (Figure 4). Unsworth and Mitchell [5] demonstrated a
mechanism by which the open-chain glucose molecule goes through a series of dehydration reactions
(elimination of water molecules) and then recycles into a 5-membered ring structure. The furanic
compounds, unlike sugars such as glucose, are oil soluble and therefore are detectable.
O
CH2OH
O
CHO
C
C
C
C
C
C
C
C
2-furfural
COCH3
C
C
C
C
furfuryl alcohol
O
CHO
C
C
C
5-hydroxymethyl-2-furfural
O
CH2OH
O
CH3
CHO
C
C
C
2-acetyl furan
C
C
5-methyl-2-furfural
COMMONLY FOUND FURANIC COMPOUNDS
FIGURE 4
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
4
ANALYSIS OF FURANIC COMPOUNDS
The method that Doble employs to perform furanic compound analysis is ASTM Method D 5837 or IEC
Method 61198. The method involves extracting semi-polar compounds, of which furanic compounds
might be present, from the oil into a more polar solvent that is then injected into a high performance liquid
chromatograph (HPLC). In the HPLC, the extracted eluent is passed through an analytical column that
separates the compounds by polarity and molecular weight. The compounds are detected by the use of a
tunable ultra-violet (UV) detector optimized at specific wavelengths for the desired compounds. Furanic
compounds are measured in ug/L. Although there are many furanic compounds, many of them are unstable
in the oil or the eluent and are of little use for routine diagnostic information. The following five furanic
compounds are commonly used for diagnostic purposes:
Common Name
5-hydroxymethyl-2-furfural
furfuryl alcohol
2-furfural
2-acetyl furan
5-methyl-2-furfural
Other Names
5-(hydroxymethyl)furfural, 5-hydroxymethyl-2-furaldehyde
2-furanmethanol, 2-furylcarbinol, furfuralcohol, 2-furfurol
2-furaldehyde, 2-furancarboxaldehyde, 2-furfuraldehyde
2-furylmethylketone
5-methylfurfural, 5-methyl-2-furaldehyde
Abbreviation
HMF
FOL
FAL
AF
MF
GENERATION, DEGRADATION, AND ACCUMULATION OF FURANIC COMPOUNDS
Furanic compounds are not completely stable in transformers and the final concentration is influenced by
numerous factors. It has been our experience that there is a constant generation and degradation of furanic
compounds so any net increase over time is considered to be accumulation.
Furanic compounds in transformers are generated from the degradation of cellulosic materials. However,
there can be residual furanic compounds in some new oils, from excessive dry-outs, or from reused oils that
should be considered in determining baseline values.
In the manufacture of new transformer oil several different refining processes can be used. One process
involves solvent refining using 2-furfural. If it is not completely removed after the refining process then it
will be present in the final product that is delivered. Residual concentrations as high as 380 ug/L have been
found by testing performed at Doble.
Used oil from electrical apparatus may also contribute furanic compounds. Many utilities use oil storage
farms in which oil from in-service apparatus is stored for reuse. This oil will often contain furanic
compounds. If this oil is not processed prior to re-use it will contaminate the next apparatus that it used in.
As shown later in this paper, reclamation of the oil is a very effective process of removing the furanic
compounds.
The treatment of the transformer and the oil at the transformer-manufacturing site might also contribute
furanic compounds. Inadvertent overheating of the cellulose during dryout and/or heat run testing might
degrade the insulation causing the production of furanic compounds. Even if the oil is removed prior to
shipment and replaced with new oil upon receipt, furanic compounds in the cellulose will partition from the
paper to the oil.
Observed General Patterns
Figures 5-9 depcit three general patterns that are often seen in transformers that have not undergone any
internal maintenance activities that would alter the amount of furanic compounds present. These examples
are taken from free breathing conservator transformer with Kraft paper insulation and sealed transformers
with thermally upgraded (TU) insulation. Figures 5 and 6 represent transformers in which the generation is
greater than the degradation of furanic compounds, thus a net accumulation occurred suggesting active oncellulose degradation.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
5
500
1200
FOL
423
400
AF
386
MF
800
796
641
600
452
450
1042
FAL
1000
Concentration, ug/L (ppb)
Concentration of Furanic Compounds, ug/L
HMF
400
362
350
376
HMF
342
322
300
298
FOL
298
FAL
277
282
AF
250
MF
200
150
100
200
50
38
0
6/19/97
1/5/98
7/24/98
2/9/99
8/28/99
3/15/00
10/1/00
4/19/01
0
7/24/1998
12/6/1999
4/19/2001
Sample Date
9/1/2002
1/14/2004
5/28/2005
10/10/2006
Sample Date
FURANIC COMPOUNDS INCREASE IN
SEALED CONSERVATOR
FIGURE 6
FURANIC COMPOUNDS INCREASE IN
OPEN CONSERVATOR
FIGURE 5
Figures 7 and 8 are graphs that indicate that the concentration of furanic compounds is relatively stable
indicating generation and degradation is nearly equal and the generation rate is low.
2500
20
HMF
2316
18
FOL
2223
2000
FAL
16
AF
Concentration, ug/L (ppb)
Furanic Compound Concentration, ug/L
2323
HMF
FOL
1500
FAL
AF
MF
1000
500
14
MF
12
10
8
6
4
2
0
6/19/97
9/27/97
1/5/98
4/15/98
7/24/98
11/1/98
2/9/99
5/20/99
8/28/99
Sample Date
STABLE CONCENTRATION OF FURANIC
COMPOUNDS IN OPEN CONSERVATOR
FIGURE 7
0
3/15/00
10/1/00
4/19/01
11/5/01
5/24/02 12/10/02 6/28/03
1/14/04
8/1/04
2/17/05
9/5/05
Sample Date
STABLE CONCENTRATION OF FURANIC
COMPOUNDS IN SEALED CONSERVATOR
FIGURE 8
Figure 9 exhibits changes in furanic compounds resulting from a very interesting history for this
transformer. In the beginning there is a rapid increase of furanic compounds, definitely suggesting there is
an active overheating condition causing accelerated degradation of the cellulose insulation. Once the 2furfural concentration reaches a maximum concentration of 3474 ug/L it starts to decline over time
indicating that the overheating condition has somehow been repaired or mitigated and the degradation of
the furanic compounds is quicker than generation, with the net effect of declining concentrations. Because
this is a free breathing conservator transformer, the reduction in furanic compounds is most likely
accelerated by the presence of oxygen. It is interesting that this type of rapid decrease is more typical in
free breathing conservator transformers then in sealed conservator transformers.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
6
3/24/06
800
4000
FAL
HMF
3474
700
FOL
AF
FAL Concentration, ug/L
3000
600
MF
2896
500
2500
400
2000
1750
1446 300
1500
1403
200
1000
813
HMF, FOL, AF and MF Concentration, ug/L
3500
100
500
0
0
10/28/95
5/15/96
12/1/96
6/19/97
1/5/98
7/24/98
2/9/99
8/28/99
3/15/00
10/1/00
Sample Date
FURANIC COMPOUNDS CHANGE OVER TIME
FIGURE 9
The amount of degradation of cellulose is based on the highest concentration of 2-furfural determined over
time not necessarily the most recent sample (Figure 9). A reduction in 2-furfural is not an indication of a
self-healing process. The reduction of furanic compounds is due to their degradation. Using the
information from Figure 9 and the conversion to DP by the Chendong equation [6], the information in
Table 1 is provided.
TABLE 1
CONVERSION OF FURANIC COMPOUNDS TO DP
2-Furfural Content, Estimated DP
ug/L
813
454
2896
297
3473
274
1750
359
1403
387
1446
383
Paper insulation with DP values below 200 are considered near or at the end of the reliable life, as it
becomes mechanically compromised. Midlife is approximately a DP of 400 and new paper typically has a
DP of about 1000 to 1300. In the example presented in Table 1, there is a large difference between the
values. In this case, the maximum concentration of 3473 ug/L of 2-furfural with an estimated DP of 274
should be used for the reasons stated previously.
Understanding the Xue Chendong Equation [6] and Its Limitations
DP is an intrusive and a destructive test as it requires a piece of paper from the transformer for analysis.
Many laboratories performing furanic compound analysis convert the 2-furfural concentration to an
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
7
estimated DP using the Chendong equation [6]. Although a powerful tool fot transformer diagnostics, it
can be misapplied. The equation is given as:
Equation 1: Estimated DP = ((log of 2-furfural in mg/kg) – 1.5)/-0.0035
It should be noted that a major component of the equation is the 2-furfural concentration. The 2-furfural
values can change due to such processes as oil change-out, vacuum processing, filtering and reclamation.
These processes will decrease the 2-furfural concentration and thus increase the estimated DP. Unless a
cumulative value was determined this can be misleading. It is always of major importance to know what
types of maintenance activities have been conducted on the transformer and the oil so that the 2-furfural
content and any diagnosis resulting from it can be accurately assessed.
The specific population of transformers that this equation is derived from were Kraft-paper wound, free
breathing conservator type transformers. There are varied practices used worldwide to manufacture
transformers. Many transformers in Latin America, Europe, and parts of Asia for example, have historically
been manufactured with breathing conservator systems and Kraft paper insulation. Most open conservator
systems usually only seal out water but not oxygen. In contrast, the most common practice in the United
States is to seal the oil and paper from air by either using a sealed tank with nitrogen or a conservator tank
with a polymer membrane. For older U.S. designs with 55°C average winding-temperature rise, Kraft
paper insulation was typically used. Most modern U.S. designs are of the 65°C average windingtemperature rise type and thermally-upgraded (TU) paper is used to wrap conductors. In some cases,
especially in mobile transformers, Nomex® insulation is also employed which is a synthetic aramide fiber
that does not produce furanic compounds.
The Chendong equation has come to be used by some to evaluate the paper DP for all transformers,
regardless of paper type or preservation system. As transformers with different insulation and preservation
systems will accumulate different amounts and ratios of furanic compounds they will have differing aging
profiles. The results of the calculation to estimated DP from the 2-furfural content with consideration of
the kind of paper and the type of preservation system. For example, this calculation when used for nitrogen
blanketed and sealed conservator transformers can seriously underestimate the degree of long-term aging
especially in transformers with thermally upgraded insulation. Under these conditions, the result could
provide a false sense of security.
Paper aging in transformers is not uniform, and follows thermal and moisture gradients. In addition,
external layers are more exposed to higher concentrations of oxygen and the byproducts of aging in the oil.
Actual DP measurements from various areas of the transformer can be drastically different depending on
the local conditions. The Chendong equation is best used to estimate an average DP in transformers with
Kraft paper insulation and free-breathing conservators. For other insulation and preservation systems the
equation can be used to estimate paper degradation from thermal events. Problems involving localized
overheating of the paper insulation will also generate furanic compounds above that from normal aging.
FACTORS THAT INFLUENCE THE CONCENTRATION AND STABILITY OF FURANIC
COMPOUNDS
The following issues have been explored in regards to the stability of furanic compounds:
•
•
•
•
Effects of oil degassing by partial vacuum
Effects of mechanical filtration (reconditioning)
Effects of oil reclamation
Effects of oil change-out
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
8
•
•
•
•
•
•
Stability of furanic compounds in oil
Kraft, thermally-upgraded (TU) Kraft, varnish and epoxy coatings and Nomex® insulation systems
The Generation of 2-Furfural in Kraft, TU and mixed insulation including the effects of dicyandiamide
(DICY) and “Cross Pollination”
Partition coefficients
Effects of electrical discharge and high temperatures
The role of the transformer preservation system
Effects of Oil Degassing by Partial Vacuum
The concentration of furanic compounds can be affected by maintenance activities on the transformer.
Applying partial vacuum to the oil to dehydrate and degas it has a collateral effect of removing furanic
compounds as well as water and gases. In a manner similar to removal of dissolved gases, certain furanic
compounds are removed in larger percentages than others. The experiments involved applying different
partial vacuum levels, temperatures and a series of passes of oil through the vacuum system. This testing
was only performed on oil and not a transformer. Table 2 and Figure 10 indicates the reduction in
concentration in each of the five furanic compounds tested.
TABLE 2
REDUCTION CONCENTATION (ug/L) OF FURANIC COMPOUNDS BY
SEVERE PARTIAL VACUUM AT 25°C (220-250 MICRONS)
COMPOUND
Initial
After 1st
After 2nd After 3rd
Concentration
Pass
Pass
Pass
5-hydroxymethyl-2-furfural
1000
914
849
803
furfuryl alcohol
1000
815
733
601
2-furfural
1000
900
729
587
acetyl furan
1000
886
838
744
5-methyl-2-furfural
1000
912
876
808
Initial Blanking Pressure: 50-70 microns
45
1200
After 1st Pass
40
Overall After 2nd Pass
1000
Overall After 3rd Pass
800
600
HMF
400
Percentage Removed
Concentration, ug/l (ppb)
35
30
25
20
15
FOL
FAL
200
10
AF
MF
5
0
0
Initial
Pass 1
Pass 2
Test Condition
FIGURE 10
Pass 3
HMF
FOL
FAL
AF
MF
Furanic Compound
FIGURE 11
Table 3 and Figure 11 indicates the reduction in concentration by percentage in each of the five furanic
compounds tested.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
9
TABLE 3
PERCENT (%) DECREASE IN CONCENTRATION FROM TABLE 2
COMPOUND
After 1st
Pass
5-hydroxymethyl-2-furfural
furfuryl alcohol
2-furfural
acetyl furan
5-methyl-2-furfural
8.6
18.5
10.0
11.4
8.8
After 2nd
Pass,
Overall
15.1
26.7
27.1
16.2
12.4
After 2nd
Pass, From
1st pass
7.1
10.1
19.0
5.4
3.9
After 3rd
Pass,
Overall
19.7
39.9
41.3
25.6
19.2
After 3rd
Pass, From
2nd pass
5.4
18.0
19.5
11.2
7.8
It was once thought that partial vacuum did not seriously affect the concentration of furanic compounds.
However, on an initial pass through a vacuum system at only 25°C and a partial vacuum level of 230
microns, of the five furanic compounds tested, an average of 10% was removed except for furfuryl alcohol
which was close to 20%. After 3 passes, factors such as solubility and molecular weight play a more
significant role with 2-furfural being reduced by more than 40% and HMF and MF being reduced by about
20% (Figure 11).
The same set of experiments were again performed at 25°C but in this case the partial vacuum level was
much less severe and ranged between 800 and 1000 microns. Table 4 and Figure 12 indicates the reduction
in concentration in each of the five furanic compounds tested.
TABLE 4
REDUCTION CONCENTRATION (ug/L) OF FURANIC COMPOUNDS BY LESS SEVERE
PARTIAL VACUUM AT 25°C (800-1000 MICRONS)
COMPOUND
Initial
Concentration
5-hydroxymethyl-2-furfural
1000
furfuryl alcohol
1000
2-furfural
1000
acetyl furan
1000
5-methyl-2-furfural
1000
Initial Blanking Pressure: 800-850 microns
1200
30
1000
25
After 1st
Pass
926
923
894
920
944
After 3rd
Pass
889
739
740
836
873
After 1st Pass
Percentage Removed
Concentration, ug/L (ppb)
Overall After 3rd Pass
800
HMF
600
FOL
FAL
400
AF
20
15
10
MF
5
200
0
0
Initial
Pass 1
Test Condition
FIGURE 12
Pass 3
HMF
FOL
FAL
AF
MF
Furanic Compound
FIGURE 13
Table 5 and Figure 13 indicates the reduction in concentration by percentage in each of the five furanic
compounds tested.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
10
TABLE 5
PERCENT (%) DECREASE IN CONCENTRATION FROM TABLE 4
COMPOUND
After 1st
Pass
5-hydroxymethyl-2-furfural
furfuryl alcohol
2-furfural
acetyl furan
5-methyl-2-furfural
7.4
7.7
10.6
8.0
5.6
After 3rd
Pass,
Overall
11.1
26.1
26.0
16.4
12.7
After 3rd
Pass, From
1st Pass
4.0
19.9
17.2
9.1
7.5
The higher the pressure (800-1000 microns), the less furanic compounds are removed but substantial
removal still occurs. After 1 pass, 5-10% of the concentration is removed, and after 3 passes furfuryl
alcohol and 2-furfural exhibited removal rates of 25% or greater.
The last set of experiments performed involved partial vacuum at a low pressure (around 250 microns) and
heating the oil to approximately 60-70°C as this is a common practice conducted in the field. The results of
this testing is provided in Table 6 and Figure 14.
TABLE 6
REDUCTION CONCENTRATION (ug/L) OF FURANIC COMPOUNDS BY SEVERE PARTIAL
VACUUM AT 60-70°C (AROUND 250 MICRONS)
COMPOUND
Initial
After 1st
After 2nd After 3rd
Concentration
Pass
Pass
Pass
5-hydroxymethyl-2-furfural
1000
878
781
639
furfuryl alcohol
1000
588
357
197
2-furfural
1000
650
403
234
acetyl furan
1000
740
525
370
5-methyl-2-furfural
1000
784
608
456
Initial Blanking Pressure: 50-70 microns
As shown in Tables 6 and 7 and Figures 14 and 15, this type of processing at elevated temperatures
severely reduced the concentration of all furanic compounds. Depending on how many passes the oil is
subjected to certain furanic compounds can be reduced by as much as 80%. It must be remembered that
these procedures were performed only on the oil and does not represent removal of furanic compounds
from the paper insulation.
1200
90
HMF
After 1st Pass
FOL
1000
80
AF
After 3rd Pass
70
MF
800
Percentage Removed
Concentration, ug/L(ppb)
After 2nd Pass
FAL
600
400
60
50
40
30
20
200
10
0
0
Initial
Pass 1
Pass 2
Test Condition
FIGURE 14
Pass 3
HMF
FOL
FAL
AF
MF
Furanic Compound
FIGURE 15
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
11
TABLE 7
PERCENT (%) DECREASE IN CONCENTRATION FROM TABLE 6
COMPOUND
After 1st
Pass
5-hydroxymethyl-2-furfural
furfuryl alcohol
2-furfural
acetyl furan
5-methyl-2-furfural
12.2
41.2
35.0
26.0
21.6
After 2nd
Pass,
Overall
21.9
64.3
59.7
47.5
39.2
After 2nd
Pass, From
1st pass
11.0
39.3
38.0
29.1
22.4
After 3rd
Pass,
Overall
36.1
80.3
76.6
63.0
54.4
After 3rd
Pass, From
2nd pass
18.2
44.8
41.9
29.5
25.0
Case Study 1 – Oil Degassing
A transformer with an appreciable quantity of the furanic compounds was investigated after overheating,
during which the dissolved gas-in-oil test (31,000 ppm carbon dioxide as shown in Table 8) indicated
insulation degradation.
TABLE 8
DISSOLVED GAS-IN-OIL RESULTS
Gas
Hydrogen
Methane
Ethane
Ethylene
Acetylene
Carbon Monoxide
Carbon Dioxide
*vol./vol. at NTP
Concentration
ppm*
291
60
82
138
0
115
31,000
The transformer was identified as follows:
Manufacturer:
Pennsylvania Transformer, Shell-form
MVA:
105
kV
132
Type of Preservation System:
Sealed by nitrogen blanket
Type of Cellulosic Insulation:
Assumed to be Kraft
Temperature:
Loading around 80%
Service Life:
Installed in 1956, service life was 39 years. Unit did not fail.
Oil Contamination:
High, serious oil degradation; acidity 0.22 mg KOH/g oil, Power Factor
of 1.5%.
It is interesting to note in Table 8 that the carbon monoxide value was quite low and not a good indicator of
paper overheating for this case. This unit had been alarming in the summer time when the simulated hotspot indicator exceeded 115°C. The internal investigation revealed free water on all horizontal surfaces,
and varnish had dissolved into the oil, discoloring it. In addition two bushings were replaced. The water
content of the solid insulation estimated from the water-in-oil content and equilibrium curves, indicated it
contained about 5%. The oil was in good condition and therefore the only treatment of the oil and
transformer before returning the oil to the unit involved flushing the transformer with dry degassed oil and
vacuum processing it for a 24-hour period. No solid adsorbent was used. Following the treatment to the
transformer, samples were taken periodically to determine the concentration of furanic compounds. It
appears that the handling of the oil and processing of the transformer resulted in a decrease in 2-furfural as
depicted in Figure 16.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
12
2-FURFURAL CONTENT
2250
Concentration of 2-furfural, ng/mL
Unit was vacuum processed for 24 hours
2000
2129
1750
1605
1500
1398
1244
1250
1194
1000
1056
1120
750
500
250
0
0
40
80
120
160
200
240
280
320
360
400
Time (Days)
EFFECTS OF OIL DEGASSING ON THE 2-FURFURAL CONTENT
FIGURE 16
The decrease could have been due to equilibration of the furanic compounds between the solid and liquid
insulation after losses during processing or due to degradation. The large increase in furanic compounds in
oil after about seven months suggested that there was again accelerated aging of the cellulosic insulation in
this transformer. A dissolved gas-in-oil test showed that the carbon dioxide had about doubled during this
time from the concentration to which it had re-equilibrated after the degassing.
Effects of Mechanical Filtration (Reconditioning)
A common maintenance activity for transformer oil is mechanical filtration. It is performed by passing oil
via a pump through a filter of a certain porosity. Although there are many types of filters and filtering
systems, filters with a porosity of 10, 5, 1, and 0.5 microns are often used and many are mainly composed
of cellulose. To replicate this in a laboratory, transformer oil was passed through a 0.45 micron nitrocellulose filter at 25°C using nitrogen to pressurize the oil through the filter instead of pulling a partial
vacuum as it was already determined that partial vacuum removed furanic compounds. The oil was filtered
three times with samples taken for analysis after the 1st and 3rd pass. The data from this testing is provided
in Table 9.
TABLE 9
REDUCTION CONCENTRATION (ug/L) OF FURANIC COMPOUNDS
BY MECHANICAL FILTRATION
%
After 3rd Total %
COMPOUND
Initial
After 1st
Pass
Decrease
Pass
Decrease
Concentration
5-hydroxymethyl-2-furfural
74.4
83.3
1000
256
167
furfuryl alcohol
28.2
44.3
1000
718
557
2-furfural
20.9
31.8
1000
791
682
acetyl furan
11.3
12.1
1000
887
879
5-methyl-2-furfural
10.8
10.1
1000
892
899
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
13
1200
Concentration, ug/L (ppb)
1000
800
600
HMF
FOL
FAL
400
AF
MF
200
0
Initial
Pass 1
Pass 3
Test Condition
REMOVAL OF FURANIC COMPOUNDS BY LABORATORY FILTRATION
FIGURE 17
It was surprising how much of an effect filtration had on the concentration of furanic compounds (Figure
17). The hydroxyl groups (OH) of each furanic compound are attracted to the cellulose comprising the
filter. In the case of 5-hydroxymethyl-2-furfural (Figure 4) where over 80% was lost after 3 passes, there
are two hydroxyl tails that comprise that molecule. The other four furanic compounds have only one
hydroxyl group. The difference in the reduction percentages of furfuryl alcohol and 2-furfural compared
with 2-acetyl furan and 5-methyl-2-furfural is the latter two compounds have an attached methyl group that
reduces the attraction to the cellulose.
Effects of Oil Reclamation
Oil reclamation is a common maintenance activity in the utility industry. For the most part adsorptive clay
(Fullers Earth) is used in conjunction with increased temperature to remove non-acidic polars such as
compounds with hydroxyl groups, acids, peroxides, ionic compounds and other products of degradation. It
is known that this type of process would reduce the concentration of furanic compounds in the oil but the
actual percentage reduction had not been measured. Oil that was spiked with furanic compounds was
passed through preheated clay (60-70°C) at a ratio of 0.25 pounds per gallon of oil (0.03 kg/L). The
process was performed in the presence of air and then nitrogen to determine if there was any significant
difference. The oil used in the test was new and not service-aged. It is not known if the use of service-aged
oil would have affected the removal efficiency of furanic compounds by the clay as the clay would have to
remove other compounds than just furanic compounds. The data from this testing is shown in Table 10,
Figures 18 and 19, and Table 11, Figures 20 and 21.
TABLE 10
REDUCTION CONCENTRATION (ug/L) OF FURANIC COMPOUNDS BY
CLAY TREATMENT AT 60-70°C IN THE PRESENCE OF AIR
COMPOUND
Initial
After 1st
After 2nd After 3rd
Concentration
Pass
Pass
Pass
5-hydroxymethyl-2-furfural
1000
31
14
<1
furfuryl alcohol
1000
<1
<1
<1
2-furfural
1000
34
10
4
acetyl furan
1000
34
11
3
5-methyl-2-furfural
1000
6
1
<1
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
14
1200
100
HMF
FOL
1000
98
FAL
96
MF
Percentage Removed
800
600
400
94
92
90
88
After 1st Pass
86
After 2nd Pass
84
200
After 3rd Pass
82
0
80
Initial
Pass 1
Pass 2
Pass 3
HMF
FOL
FAL
Test Condition
Furanic Compound
FIGURE 18
FIGURE 19
AF
MF
TABLE 11
REDUCTION CONCENTRATION (ug/L) OF FURANIC COMPOUNDS BY
CLAY TREATMENT AT 60-70°C IN THE PRESENCE OF NITROGEN
COMPOUND
Initial
After 1st
After 2nd After 3rd
Concentration
Pass
Pass
Pass
5-hydroxymethyl-2-furfural
1000
15
12
10
furfuryl alcohol
1000
65
8
<1
2-furfural
1000
138
35
11
acetyl furan
1000
118
24
5
5-methyl-2-furfural
1000
85
12
2
1200
100
HMF
98
FAL
AF
MF
800
600
400
Percentage Removed, ug/L (ppb)
FOL
1000
Concentration, ug/L (ppb)
Concentration, ug/L (ppb)
AF
96
94
92
90
88
86
After 1st Pass
84
200
After 2nd Pass
82
0
After 3rd Pass
80
Initial
Pass 1
Pass 2
Test Condition
FIGURE 20
Pass 3
HMF
FOL
FAL
AF
MF
Furanic Compound
FIGURE 21
It is believed that the clay processing performed under nitrogen would better replicate field processing as
the oil usually is pumped through the Fuller’s earth towers and thus would not necessarily be saturated with
air. Although the reduction of furanic compounds is significant in both cases, furanic compounds in oil that
was air saturated had a greater than 98% extraction efficiency for the compounds tested.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
15
Case Study 2 – Oil Reclamation
A large shell-form transformer with some gassing problems was going through repair. In that process the
oil from the unit was subjected to clay treatment. The oil reclamation involved the use of twin 600 pound
clay towers to process 9500 gallons of oil. The transformer was identified as follows:
Manufacturer:
MVA:
kV
Type of Preservation System:
Type of Cellulosic Insulation:
Volume:
Manufacturer Year
Westinghouse, Shell-form
333
500
Sealed Conservator System
Assumed to be Thermally-upgraded (TU) Kraft
9400 gallons of mineral oil
1971
Table 12 exhibits the furanic compound data before and after the oil reclamation process. The transformer
was in service but just for a few weeks when the “After Clay Treatment” sample was taken and
equilibration of the furanic compounds between the oil and paper most likely had not been reached.
TABLE 12
REDUCTION CONCENTRATION (ug/L) OF FURANIC COMPOUNDS
BY CLAY TREATMENT ON A SHELL-FORM TRANSFORMER
Test Date
HMF
FOL
FAL
AF
Before Clay Treatment 12/12/2005
<1
<1
18
1
After Clay Treatment
1/8/2006
<1
<1
3
<1
MF
8
<1
Case Study 3 – Oil Reclamation
This was a core form unit that was showing signs of overheating of the cellulose insulation as indicated by
the elevated carbon dioxide concentration in the DGA analysis (12,600 ppm CO2, and 623 ppm CO). The
DGA results also indicated some low temperature overheating of the oil was taking place. The transformer
was identified as follows:
Manufacturer:
MVA:
kV
Type of Preservation System:
Type of Cellulosic Insulation:
Volume:
Manufacturer Year
Westinghouse, Core-form
56
115
Gas Blanketed System
Assumed to be Thermally-upgraded (TU) Kraft
14400 gallons of mineral oil
1986
The quality of the oil in this unit was in very poor condition. The neutralization number and power factor at
100°C was very elevated at 0.32 mg KOH/g and 9.03% respectively and the interfacial tension being very
low at 18 mN/m thus indicating severe oil degradation had occurred. In addition, a Doble sludge tested
performed and indicated the presence of soluble sludge. There was also an indication that the unit was
fairly wet. As a result, the utility decided to perform oil processing on the oil in the unit. The results of the
furanic compound testing performed before and after oil processing is provided in Table 13. The oil
quality results are provided in Table 14. It is not known how long the transformer had been energized after
the oil reconditioning took place and the “After Clay Treatment” sample was taken. Therefore the time is
also unknown.
TABLE 13
CONCENTRATION (UG/L) REDUCTION OF FURANIC COMPOUNDS
BY CLAY TREATMENT ON A SHELL-FORM TRANSFORMER
Test Date
HMF
FOL
FAL
AF
Before Clay Treatment
6/4/2004
<1
<1
613
1
After Clay Treatment
8/10/2005
<1
<1
102
1
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
MF
35
4
16
TABLE 14
OIL QUALITY DATA BEFORE AND AFTER OIL PROCESSING
Test Date
Neut. No.
IFT
PF at 25°C, PF at 100°C,
mgKOH/g
mN/m
%
%
Before Clay Treatment
4/9/2004
0.32
18
0.316
9.03
After Clay Treatment
8/10/2005
0.01
30
0.033
1.616
The oil processing work that was performed was successful in returning the oil condition to an acceptable
state. In both cases, there was also a significant removal of furanic compounds with 2-furfural being
reduced by 83% (Case 2 and 3) and 5-methyl-2-furfural being reduced by 89% (Case 3). Both values are
extremely close to the experimental values for a single pass in Figure 21. Both showed that clay treatment
had similar effects on removal of furanic compounds. In neither case has there been additional furanic
compound tests
Effects of an Oil Change-out
Laboratory experiments were not performed to evaluate this particulate oil-processing procedure. Instead,
an actual case is used to illustrate how the furanic compounds are affected after an oil change-out.
Case Study 4 – Oil Change-Out
This particular utility was taking some of their small silicone-filled transformers, draining and flushing the
silicone out of the transformer and replacing it with mineral oil. They performed regular furanic compound
analysis on this unit before and after the oil change-out so the history is well documented in Figure 22. The
transformer was identified as follows:
Manufacturer:
MVA:
kV
Type of Preservation System:
Type of Cellulosic Insulation:
Volume:
Manufacturer Year
Sola Basic Industries
0.5
4.2
Gas Blanketed System
Unknown
440 gallons of silicone changed to mineral oil
1974
100
90
Concentration, Ug/L (ppb)
80
HMF
Transformer drained and
flushed of silicone and mineral
oil installed
70
FOL
FAL
AF
60
MF
50
Top Oil, C
40
30
20
10
0
03/15/00 10/01/00 04/19/01 11/05/01 05/24/02 12/10/02 06/28/03 01/14/04 08/01/04 02/17/05 09/05/05 03/24/06
Sample Date
FIGURE 22
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
17
The reduction of 2-furfural and other compounds was significant as a result of the oil change-out with 2furfural being reduced from 62 to 9 ug/L (85% reduction). It has since increased to 17 ug/L which is still
73% lower than the original value since the oil change-out occurred in 2001. This likely represents
generation. It suggests that removal of a good portion of the furanic compounds in the transformer system
occurred during the retrofill.
Stability of Furanic Compounds in Oil
The rate of accumulation of some furanic compounds in oil increases with increasing temperature. As
indicated previously, the term accumulation (the sum of furanic compound generation and degradation) is
used because it has been shown that furanic compounds can be unstable at 120°C under some conditions
[7,8]. The results on the rate of degradation of the furanic compounds can be summarized as follows:
Rate of Furanic Compound Degradation
Condition
SLOW
MODERATE
RAPID
Air (Oxygen)
Copper
DICY
Copper & DICY
Copper & Air (oxygen)
Nitrogen
DICY = dicyanodiamide (one compound used as a thermally upgrading agent in thermally upgraded paper)
The stability of furanic compounds is of interest because investigators have attempted to use the total
amount of these compounds or a specific one to estimate the overall aging of cellulosic materials. The
accumulation of furanic compounds is dependent upon competing reactions. Large amounts of the furanic
compounds can be generated when cellulosic materials are exposed to temperatures above 120°C. The rate
is dependent upon the type of insulation and the conditions to which it is exposed, such as temperature,
water content, and concentration of oxygen. Once formed, these furanic compounds can then survive for
prolonged periods of time in the bulk oil, which is at a significantly lower temperature than the hottest-spot
insulation. Further testing has shown that furfuryl alcohol is unstable under a number of conditions and
therefore is not very useful for diagnostics. It has been our experience that the most common furanic
compound detected is 2-furfural (FAL). It is found in most samples from transformers. 5-methyl-2furfural has been the second most common furanic compound found in our results Currently, most of the
diagnostics using furanic compound analysis is based on the concentration of 2-furfural. However, to use
furanic compounds for assessing the condition of the cellulosic insulation the type of solid insulation needs
to be considered.
Kraft, Thermally-Upgraded (TU) Kraft, Varnish and Epoxy Coatings and Nomex® Insulation Systems
Distinct differences in concentration and relative composition of furanic compounds can be found in
transformers that are composed of different types of insulation systems. The oil expansion preservation
system is also a factor.
Kraft paper insulation systems are made of electrical grade Kraft paper with no additional chemicals added.
Kraft paper is primarily composed of cellulose with linear and polymeric chains of repeating glucose units.
It also contains small amounts of hemicellulose and lignins [9] of which the actual amount depends on the
pulping process. Kraft paper insulation systems are commonly used in many parts of the world until
recently. This type of system, in general, will produce more furanic compounds under the same conditions
than TU-Kraft paper systems.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
18
TU-Kraft paper insulation systems are composed of two general types used for electrical apparatus. One
type is manufactured using a cyanoethylation process and was developed by General Electric with the trade
name of Permalex® [10]. Since it was known that Kraft paper was very strong mechanically, but thermally
deteriorated, efforts to make the paper more thermally stable were attempted. General Electric developed
the process in the late 1950s/early 1960s. Cyanoethylation resulted in a change of the chemical structure of
the paper by substituting a cyano group containing nitrogen atoms for hydrogen atoms (Figure 23). As a
result of the reaction with the primary hydroxyl group, this effectively physically and chemically hindered
degradation reactions.
CYANOETHYLATED CELLULOSE
FIGURE 23
In another process, a different family of nitrogen-bearing chemicals were added to the paper to provide
thermal stability. The earliest practice was the use of urea that was hung in a bag inside the transformer. In
the late 1950s/early 1960s Westinghouse developed a chemical addition process which is now the most
prevalent, as it ended up being the least expensive to make. Insuldur, was Westinghouse’s trade name for
TU-Kraft paper. It consisted of electrical grade Kraft paper treated with the chemical compounds
dicyandiamide, melamine and polyacrylamide in the 2-4% range by weight. These chemicals, which are
applied directly to the paper, hydrogen bonds to the paper molecule but does not structurally alter it.
It is hard to determine the effects of the General Electric cyanoethylation process on the generation of
furanic compounds because it is no longer a process that is used and thus no virgin samples exist on which
to perform experiments.
Varnishes and epoxy-coated systems do not produce furanic compounds as there is no cellulose in these
systems. The same is true for Nomex® wound conductor as it is a synthetic fiber and not composed of
cellulose from which furanic compounds are generated. In these cases furanic compounds cannot be used
as indicators of insulation breakdown. In insulation systems that are mixed such as varnish/cellulose and
Nomex®/cellulose, furanic compounds will be produced. It is expected that the concentrations will be
lower as the amount of cellulose insulation present is reduced.
The Generation of 2-Furfural in Kraft, TU and Mixed Insulation including the Effects of DICY and “Cross
Pollination”
The amount and type of furanic compounds generated is different for Kraft and TU-Kraft paper. A
comparison of the relative composition of byproducts formed under the same controlled laboratory
conditions is given in Table 15.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
19
TABLE 15
PERCENT COMPOSITION OF PAPER BYPRODUCTS IN OIL
Compound
5-Hydroxymethyl-2-furfural (HMF)
Furfuryl alcohol (FOL)
2-Furfural (FAL)
Acetyl furan (AF)
5-Methyl-2-furfural (MF)
Kraft
23.3
0.1
60.4
2.0
14.2
TU-Kraft
0.4
72.5
19.8
4.1
3.3
Studies at Doble have consistently shown 90-95% of samples from transformers with thermally upgraded
insulation to have less than about 100 ug/L of 2-furfural. This applies to both mineral oil and silicone
transformers [11,12]. However, for a similar population study of oils from service-aged transformers with
Kraft paper insulation, the 95% norm was 2057 μg/L [13]. Part of the reason for this difference can be seen
from laboratory studies of paper aging comparing Kraft, TU Kraft, and mixed insulation. Figure 24 reveals
the results of the study showing the different patterns for 2-furfural accumulation. This is described below.
•
•
•
Kraft paper - there is a continuously increasing value of FAL
TU Kraft - FAL increases initially and then decreases,
Mixed insulation system of Kraft and TU Kraft- essentially behaves as a TU system after an initial
higher increase in its content.
2-FURFURAL ACCUMULATION
120
Kraft
MIXED
TU
2-FURFURAL CONTENT, ug/L
100
80
60
40
20
0
0
5
10
15
20
25
30
35
40
45
50
55
60
TIME (Days)
2-FURFURAL GENERATION WITH DIFFERENT PAPERS
FIGURE 24
As shown in Figure 24, Kraft paper insulation produces the highest concentrations of 2-furfural. The
mixed insulation system (TU-Kraft and Kraft), typical of many U.S. transformers, yielded concentrations
which were between the two but closer to the TU-Kraft results. The results showed that there is a "crosspollination" effect where the aging of the Kraft paper was retarded as determined by DP measurements
when it was aged in the presence of DICY type TU-Kraft paper. For example, experiments where the DP
was very low at 300-450, the amount of furanic compounds present were also very low suggesting there
was little aging of the paper. This suggests that the thermally upgrading agents used retards furanic
compound production or reacts with them. With thermally-upgraded insulation systems furanic compounds
are not always accurate indicators of cellulose degradation. Certain furanic compounds such as 5© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
20
hydroxymethyl-2-furfural would appear to be of limited value as it disappears in thermally-upgraded Kraft
systems. The other furanic compounds, 2-furfural, 5-methyl-2-furfural, and acetyl furan appear to be
reasonably stable. A change in concentration of these compounds is indicative of an increase in insulation
temperature or its aging rate and is extremely useful in this respect.
Partition Coefficients
The furanic compounds will partition between the oil and the paper. The partition coefficients may be
dependent on temperature and oil quality [7], but further research is needed to confirm these findings.
They should be independent on the amount of paper present. Other factors such as the amount of thermally
upgrading agents found in shell-form versus core-form transformers may be the reason why these
transformer designs exhibit different furanic compound concentrations when treated as separate
populations.
Effects of Electrical Discharge and High Temperatures
In Doble studies it was found that the furanic compounds are not detected in high quantities when cellulosic
materials are subjected to electrical discharges. Even though the paper insulation was significantly
damaged and carbon oxides formed, the concentration of furanic compounds was very low. Therefore, the
energy produced by the electrical discharge creates a high enough temperature in the local vicinity to
degrade furanic compounds as soon as they are formed [8]. This strongly suggests that furanic compounds
have a high-temperature instability. This information has been very valuable to help distinguish between
thermal events resulting in failure and just dielectric failures.
The Role of the Transformer Preservation Systems
Although there are many different variations on a theme, there are two basic types of transformer
preservation systems. One type is considered to be restricted breathing where the oil is open to the outside
air in a conservator expansion tank. Drycol and silica gel breathers are used to remove moisture from the
incoming air but oxygen is not excluded from the system. The second type is a sealed system where
basically the inside of the transformer is sealed from the outside environment. Types of sealed systems
include:
•
•
•
•
Nitrogen gas blanket in the headspace of a transformer
Hermetically sealed transformers with expandable bellows
Constant pressure nitrogen system where if nitrogen is lost from the transformer it is replenished from
a nitrogen bottle or generator
Bladder or diaphragm in a conservator tank that allows expansion and contraction of the oil
Sealed systems, if maintained properly, exclude moisture and oxygen from the insulation. Reducing the
amount of water and oxygen in the transformer effectively reduces the aging of the cellulose and thus the
production of furanic compounds.
Suggested Limits, Trends and Rates of Accumulation of Furanic Compounds
Furanic compounds can be significantly degraded by high temperatures, particularly in the presence of high
oxygen, and appear to react with the DICY used to thermally upgrade Kraft paper. For these reasons, Doble
instituted a diagnostic scheme with regards to four types of transformer systems as follows:
•
•
•
•
Kraft paper insulation and free breathing conservator
Kraft paper insulation and sealed system
TU-Kraft paper (chemical addition) and free breathing conservator
TU-Kraft paper (chemical addition) and sealed system
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
21
The highest values of furanic compounds are found for service-aged Kraft paper insulation in transformers
with free-breathing conservators. For sealed systems with the DICY thermally-upgraded insulation, the
amounts of furanic compounds appear to remain fairly constant unless there is a thermal event or a thermal
incipient-fault condition, in which case there is usually a step increase. In thermally-upgraded transformer
insulation containing DICY, 5-hydroxymethyl-2-furfural would also appear to be of limited value as it
disappears under these circumstances. This may be equally true for free-breathing conservator systems
with high oxygen contents in the oil, regardless of paper type. The other furanic compounds, 2-furfural, 5methyl-2-furfural, and acetyl furan appear to be reasonably stable in sealed transformers and to a lesser
extent in free-breathing conservator transformers.
Based on our experience we have developed limits that are useful to diagnosis cellulose degradation. The
accumulation rate of 2-furfural is used extensively in our analysis as shown in Table 16.
TABLE 16
2-FURFURAL ACCUMULATION RATES
Transformer Type
Accelerated Aging
Kraft Paper/Free Breathing
≥50 ug/L/year
Kraft Paper/Sealed
≥25 ug/L/year
TU-Kraft paper/Free Breathing
≥35 ug/L/year
TU- Kraft paper/Sealed
≥20 ug/Lyear
Additional Guidelines are as follows:
For Kraft paper insulation in Free Breathing Conservator Systems our guidelines are as follows:
•
The Chendong equation can be used to provide a general sense of the average DP based on the 2furfural content. However, it is always best to have some DP data for a given population to verify the
relationship.
•
Values of 2-furfural >1000 ug/L should raise a flag for further study and close monitoring
•
If estimating the DP from the 2-furfural content using the Chendong equation the guideline should be:
•
•
•
Estimated DP >800
Estimated DP <800 > 400
Estimated DP <400
Cellulosic insulation is in Good condition
Cellulosic insulation is at Midlife
Cellulosic insulation is in Last third of life
For Kraft paper insulation in Sealed Systems (Sealed, Gas Blanketed, Sealed Conservator), our guidelines
are as follows:
•
The Chendong equation can be used to provide a very general sense of the average DP based on the 2furfural content. However, rely more heavily on the 2-furfural accumulation rate.
For TU paper insulation using the dicyandiamide process in Free Breathing systems our guidelines are as
follows:
•
If estimating insulation quality from the 2-furfural content, use these guidelines:
• <100 ug/L
Normal aging if transformer is greater than 5 years old
• > 100 <700 ug/L
Midlife (examine rate)
• > 700 ug/L
possibly last third of life, unit should be monitored closely
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
22
For TU paper insulation using the dicyandiamide process in Sealed Systems (Sealed, Gas Blanketed,
Sealed Conservator), our guidelines are as follows:
•
If estimating insulation quality from the 2-furfural content, use these guidelines:
• <100 ug/L
Normal aging if transformer is greater than 5 years old
• > 100 <500 ug/L
Midlife (examine rate)
• > 500 ug/L
possibly last third of life, unit should be monitored closely
Examination of the trend is very important to distinguish between long-term aging and thermal events.
Precautions
It should be kept in mind that there is not a method to tell if the byproducts such as the carbon oxide gases
and furanic compounds are from a small amount of insulation heated to a very high temperature or a larger
mass of insulation at a lower temperature. It is therefore difficult to predict the general aging of the
insulation system based on quantities of these compounds. Severely damaged areas may not be adequately
identified because the small amount of insulation involved. Equilibration of furanic compounds between
the paper and oil takes at least a few months. Another sample should be taken after three months to
determine baseline conditions. This is also true for maintenance activities. Samples for furanic compounds
should be taken about three months after any processing of the oil or transformer.
TRANSFORMER POPULATIONS
It is useful to develop typical values of furanic compounds for a population of transformers. Ideally this
would be related to DP values on paper samples taken when opportunities present themselves such as when
internal investigations are performed.
Comparison of sister transformers within a substation that are subjected to the same load and environmental
conditions provides interesting and sometimes critical information. In Table 17 all the units are generating
basically the same amount of furanic compounds at a very low level. These gas-blanketed transformers
were built in 1951 and contain 4410 gallons of oil.
S/N
…261
…262
…263
…264
…266
TABLE 17
FURANIC COMPOUNDS IN SISTER TRANSFORMERS
Manuf.
MVA Kv
Year
Sub
HMF
FOL
FAL
Built
Pennsylvania
29
230
1951
A
<1
<1
3
Pennsylvania
29
230
1951
A
<1
<1
6
Pennsylvania
29
230
1951
A
<1
<1
4
Pennsylvania
29
230
1951
A
<1
<1
5
Pennsylvania
29
230
1951
A
<1
<1
5
AF
MF
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
Table 18 shows two sister transformers in the same substation that have similar furanic compound levels
but at a much higher concentration then the previous population. Even though the cellulose insulation is
aging at an accelerated rate, the design and use of the transformers cause them to be aged in a similar
manner.
S/N
…337
…338
TABLE 18
FURANIC COMPOUNDS IN SISTER TRANSFORMERS
Manuf.
MVA Kv
Year
Sub
HMF
FOL
FAL
Built
Westinghouse
40
21
1984
B
3
<1
784
Westinghouse
40
21
1984
B
<1
<1
740
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
AF
MF
<1
1
27
24
23
The units in Table 19 show furanic compound concentrations that are comparable between all twelve units.
Some have a little more or a little less 2-furfural than others but it is a population that represents a similar
aging pattern. Situations in which a transformer is low on oil, may have a cooler blocked, the placement
and the location subjects it to higher ambient temperatures are all reasons why the concentration can be
different. In this case, since the units were built in 1930 there is expected to be some variability. The one
unit highlighted in this group of twelve units clearly shows that it is aging differently than the other as the
concentration of 2-furfural is about double. If the unit was more modern, then some investigation may be
warranted but because of the age, there is no real reason to investigate this unit as the rate of 2-furfural is
still very low at 3.1 ug/L/year.
S/N
…7568
…7569
…7570
…7571
…7572
…7574
…7575
…7576
…7577
…7578
…7579
…7580
TABLE 19
FURANIC COMPOUNDS IN SISTER TRANSFORMERS
Manuf.
MVA Kv
Year
Sub
HMF
FOL
FAL
Built
General Electric
13
230
1930
C
<1
<1
75
General Electric
13
230
1930
C
<1
<1
85
General Electric
13
230
1930
C
<1
<1
77
General Electric
13
230
1930
C
<1
<1
76
General Electric
13
230
1930
C
<1
<1
101
General Electric
13
230
1930
C
<1
<1
66
General Electric
13
230
1930
C
<1
<1
89
General Electric
13
230
1930
C
<1
<1
100
General Electric
13
230
1930
C
<1
<1
235
General Electric
13
230
1930
C
3
<1
88
General Electric
13
230
1930
C
<1
<1
115
General Electric
13
230
1930
C
<1
<1
87
AF
MF
15
19
15
<1
17
<1
<1
<1
<1
4
16
13
8
10
12
18
10
10
2
11
21
1
21
11
The units in this substation (Table 20) were composed of station service units and GSU units. The two
Westinghouse SST units have a very similar pattern with a low production of 2-furfural. The GE GSU
units show a distinct difference between the two. Even though 90 ug/L is not excessive for a unit from
1971 it is much different than its sister. An increased sample frequency of every couple years or so may be
required to monitor this transformer.
RCR…01
RCR…11
TABLE 20
FURANIC COMPOUNDS IN TWO GROUPS OF SISTER TRANSFORMERS
Manuf.
MVA
Kv
Year
Sub
HMF
FOL
FAL
AF
Built
Westinghouse
12
13.8
1970
D
<1
<1
2
<1
Westinghouse
12
13.8
1970
D
<1
<1
4
<1
D…398
D…397
General Electric
General Electric
S/N
500
500
345
345
1971
1971
D
D
<1
1
<1
<1
14
90
MF
<1
<1
<1
1
1
8
Table 21 shows that one of the Wagner units is definitely exhibiting signs of cellulose overheating. Since
the concentrations of furanic compounds are so different and much more elevated than the sister unit an
investigation is warranted to determine the reason for the overheating.
S/N
…150
…152
TABLE 21
FURANIC COMPOUNDS IN SISTER TRANSFORMERS
HMF
FOL
FAL
Manuf.
MVA
Kv
Year
Sub
Built
Wagner
1.9
70.6
1965
E
397
35
451
Wagner
1.9
70.6
1965
E
<1
<1
9
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
AF
MF
32
<1
95
<1
24
BUSHINGS, OIL CIRCUIT BREAKERS (OCBS) AND LOAD TAP CHANGERS (LTCS)
Furanic compound analysis is not usually performed on bushings, OCBs or LTCs. However, Doble has
performed testing on all three types of oil-filled apparatus and accessories. Testing of bushings probably
hold the most promise as there is a small amount of oil compared to the amount of paper available in the
bushing thus providing a concentrating effect. In addition, most oil/paper bushings are constructed using
Kraft paper insulation that tends to yield higher concentrations of furanic compounds. Table 22 provides
some data from two bushings. As highlighted there is one data point that really stands out and indicates a
severe problem and one that might be developing a problem. More and more, utilities are sampling
bushings for DGA. Doble would recommend water content and furanic compounds as well which can all
be performed on one syringe of oil. This would provide some very good diagnostic information on the
condition of the solid and liquid insulation in the bushing. Furanic compounds data can be particularly
useful for bushings that operate warm and could have compromised seals or gaskets that allows gaskets to
escape and air to enter.
TABLE 22
FURANIC COMPOUNDS IN BUSHINGS
Manuf.
HMF
FOL
FAL
AF
MF
GE
195
45
10200
85
562
ABB
164
0
183
4
67
One would not normally think of an OCB as having enough cellulose material to produce large
concentrations of furanic compounds. However, even in the small sample population that is available,
several OCBs have shown 2-furfural concentrations over 100 ug/L indicating breakdown of possibly the
arc chutes, tank liners, wood lift rods, or some other material of construction. The concentration of furanic
compounds produced in OCBs would be predicated on the amount of cellulosic material available for
degradation.
Analysis for furanic compounds in LTCs has been very encouraging as a diagnostic tools. The production
of furanic compounds, like dissolved gases-in-oil, is most likely model specific and again predicated on the
amount of cellulose material in the LTC. A few examples of testing performed on LTCs is provided below.
TABLE 23
FURANIC COMPOUND CONCENTRATIONS IN LTCs, ug/L and CARBON DIOXIDE (ppm)
Manu
ABB
AC
Alstom
Asea
Asea
Federal Pac
Ferranti Pack
Ferranti Pack
Ferranti Pack
Ferranti Pack
Ferranti Pack
GE
McGraw Edison
Pennsylvania
Reinhausein
Westinghouse
Westinghouse
Westinghouse
Model
TLH-21
RH Mlll 3504-72.S/B
UZ BLT
UZ BLT 200500
550C
550 BLS
RMV-11
URT
URT
Type
SG Breather
FB
SG Breather
SG Breather
SG Breather
Compart
Gallons
147
150
525
525
525
525
525
150
380
190
Vacuum
FB
FB
FB
Transfer
Selector
280
263
263
HMF FOL FAL AF MF CO2
0
0
6
0
0
679
0
0
8
0
0
0
0
1
2
0 2250
0
206
0
0 4068
0
58
0
0 2750
0
0
25
0
4
0
219
0
0 8330
20
1064 5
12 6200
0
0 1196 0
0
594
0
0 1174 0
0 4481
20
0 1428 35 26 5640
0
0
28
0
2 3420
0
0
3
1
0
650
0
0
3
0
2
533
0
0
145
0
0
0
0
9
0
0 1031
0
0
6
0
2 1390
0
0
4
0
0 1880
As shown in Table 23, several LTCs have produced significant amounts of furanic compounds. Like
transformers, it is expected that the normal concentration of some models should be fairly low. It is
interesting to note that in a majority of the cases where high furanic compounds exist in high
concentrations, the carbon dioxide concentration is also elevated but this is not always true. as shown in the
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
25
several that are highlighted.
compounds.
Some LTC models appear to produce high concentrations of furanic
CONCLUSIONS
The analysis of furanic compounds in oil can provide valuable information for assessing the condition of
the cellulose insulation. It is also apparent that the more specific information known about a transformer
and its family, the better the diagnosis, as populations can be significantly different. Furanic compounds
are reduced in concentration by such maintenance activities such as oil degassing, change-out, reclaiming
and reconditioning. Examination of residual amounts of furanic compounds after maintenance activities has
been shown.
Partial discharge and arcing conditions can locally damage the cellulose insulation and yet not produce
significant quantities of furanic compounds. They appear to be degraded at the temperatures in which those
incipient-fault conditions take place.
Chemicals in the insulation used for thermal upgrading such as dicyandiamide significantly reduce the
concentration of furanic compounds in the oil.
Data from a furanic compound analysis should be treated carefully. As shown in this paper, Doble was
developed a scheme for determining the difference between normal and accelerated aging based on the
insulation type and preservation system present. The categories are as follows:
•
•
•
•
Kraft paper insulation and free breathing conservator
Kraft paper insulation and sealed system
TU-Kraft paper (chemical addition) and free breathing conservator
TU-Kraft paper (chemical addition) and sealed system
It has been found that transformers in these categories produce different amounts of furanic compounds at
varying rates and should not be grouped together as a single population. Using rates of accumulation aids
in determining if a fairly new transformer may have an abnormal overheating condition much sooner than
concentration limits alone. For example if a new transformer (sealed, TU-Kraft insulation) within the first
six months of operation produces a concentration of 35 ug/L, in most cases it would assumed to operating
normally as it is less than 100 ug/L. However, the 2-furfural accumulation rate is 70 ug/L/year which is
excessive and indicative of accelerated heating of the cellulose insulation.
References
[1] Lewand L.R. and Griffin P.J., “How to Reduce the Rate of Aging in Transformer Insulation”, NETA
World, Spring, 1995.
[2] McNutt, W. J. "Insulation Thermal Life Considerations For Transformer Loading Guides", IEEE
Trans. on Power Delivery, Vol. 7, No. 1, Jan. 1992, Pg. 392-401.
[3] Griffin, Paul J., "Measurement of Cellulose Insulation Degradation: A Study of Service-Aged
Transformers," Minutes of the Fifty-Ninth Annual International Conference of Doble Clients, 1992,
Sec. 10, pp. 4.1-4.31.
[4] Casey, James P. “Pulp and Paper; Chemistry and Chemical Technology”, Interscience Publishers,
Inc., New York, 1952.
[5] Unsworth, J. and Mitchell, F. "Degradation of Electrical Insulating Paper Monitored with High
Performance Liquid Chromatography," IEEE Transactions on Electrical Insulation, Vol. 25, No. 4,
August 1990, pg. 737-46.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
26
[6] Chendong, X. “Monitoring Paper Insulation Aging by Measuring Furfural Contents in Oil”, 7th
International Symposium on High Voltage Engineering, Aug. 26-30, 1991, pp. 139-142.
[7] Griffin, P.J., and Lewand, L.R., “A Practical Guide for Evaluating the Condition of Cellulosic
Insulation in Transformers”, Proceedings of the Sixty-Second Annual International Conference of
Doble Clients, 1995, Sec. 5-6.
[8] Griffin, P.J., Lewand, L.R., and Pahlavanpour, B., “Paper Degradation By-Products Generated Under
Incipient-Fault Conditions”, Minutes of the Sixty-First Annual International Conference of Doble
Clients, 1994, Sec. 10-5.
[9] Moser, H.P., Dahinden, V., Friederich, H., Kunast, H., Lennarz, K., Leukens, U., and Potocnik, O.,
“Transformerboard”, Scientia Electrica, 1979
[10] Raab, E. L. "Permalex--A New Insulation System for Sealed, Liquid-Immersed Apparatus," Minutes
of the Twenty-Seventh Annual International Conference of Doble Clients, Sec. 6-301, 1960.
[11] Griffin, P. J., Finnan, E., Lewand, L. R. “Case Studies”, Proceedings of the Sixty-Third Annual
International Conference of Doble Clients, 1996, Sec. 5-4.
[12] Griffin, P. J., Lewand, L. R., and Finnan, E., “Evaluation of Electric Apparatus-Case Studies”,
Proceedings of the Sixty-Fifth Annual International Conference of Doble Clients, 1998, Sec. 5-2.
[13] Finnan, E., et al., “A Report on the Assessment of Insulation Aging and Condition by Means of
Laboratory Oil Tests”, Proceedings of the Sixty-Fourth Annual International Conference of Doble
Clients, 1997, Sec. 5-5.
AUTHOR BIOGRAPHY
Lance Lewand
Lance Lewand is the Laboratory Manager for the Doble Materials Laboratory
and is also the Product Manager for the Doble DOMINO®, a moisture-in-oil
sensor. The Materials Laboratory is responsible for routine and investigative
analyses of liquid and solid dielectrics for electric apparatus. Since joining
Doble in 1992, Mr. Lewand has published numerous technical papers
pertaining to testing and sampling of electrical insulating materials and
laboratory diagnostics. Mr. Lewand was formerly Manager of Transformer
Fluid Test Laboratory and PCB and Oil Field Services at MET Electrical
Testing Company in Baltimore, MD for seven years. His years of field service
experience in this capacity provide a unique perspective, coupling laboratory
analysis and field service work. Mr. Lewand received his bachelor of science degree from St. Mary's
College of Maryland. He is actively involved in professional organizations such as ASTM D-27 since 1989
and is a sub-committee chair. He is also the secretary of the Doble Committee on Insulating Materials.
© 2006 Doble Engineering Company – 73rd Annual International Doble Client Conference
All Rights Reserved
View publication stats
27
Download
Related flashcards
Create Flashcards