Document LKkjv7nR0a585jdybw44gGbNw
t
DOBLE ENGINEERING COMPANY
ELECTRICAL INSULATION ENGINEERS AND CONSULTANTS FIELD TESTING AND MAINTENANCE OF HIGH VOLTAGE INSULATION
June 4, 1971
Mr. J. C. Wygant Monsanto Company Organic Chemicals Division 800 North Lindbergh Boulevard St. Louis, Missouri 63166
Dear Mr. Wygant:
Report No. 3 Electrical Insulating Fluid MCS 1112
Lot OR 180213
In accordance with your letter of March 11, we have per formed tests on a sample of the subject fluid which we received in Belmont on March 24, 1971. The results of our tests are included in the attached copy of cur Labo ratory Report C-1236 (Sheets A-H). For comparison pur poses, we have also listed the test limits specified in our Transformer Oil Purchase Specification (TOPS), and have included the results of tests performed on a sample of currently acceptable electrical insulating oil pro duced by sulfuric-acid refining of a predominantly naphthenic-base crude. Our tests consisted of:
Screen tests, including
1. Dielectric breakdown of samples before and after aging
2. Neutralization Number of samples vs. hours of aging
3. Interfacial Tension of samples vs. hours of aging
B. Oxidation Tests
1. Sludge-free life
2. Power-factor valued oxidation (PFVO) C. Impulse Breakdown, before and after aging D. Gassing Tendency
$ DSW 201710
32 Locust Street Belmont, Massachusetts 02178 Tel. 617-489-2830 Cable Address: DOBLENG
STLCOPCB4059777
f Addressee Monsanto Company
Date June 4, 1971
Page
2
. We believe that our comparative screen-test data speak for themselves, noting that the Aniline Point and Viscosity constitute two areas of considerable differ ence between the two samples compared.
Our Sludge-Free Life and PFVO tests are commonly applied measurements of the chemical and electrical stability of insulating oils. As predicted, the Sludge-Free Life test does not appear to be applicable to dodecyl benzene, no second phase (end point) was observed during 140 hours of aging time. The acid-refined oil sample developed .incipient sludge after 80 hours; therefore, its SludgetFree Life is recorded as 72 hours (the last sludge-free sampling period). It would appear that better bases 'for comparison of the two fluids are offered by the PFVO Characteristic (power factor versus time), and the plots of Neutralization Number and Interfacial Tension versus aging time, all of which are included in the report.
The results of Impulse Breakdown tests before and after
aging are tabulated for both fluids for comparison pur
poses.
_
<s
Gassing Tendency tests indicate that the dodecyl benzene sample is an "absorber" based on our test. The latter
is discussed in detail in the enclosed copy of a 1960
Doble Client Conference paper by Dr. G. H. von Fuchs.
We hope^tnat the foregoing comments and our enclosures
provide some of the information you require regarding
the sample of dodecyl benzene which you submitted to us.
Following your review of our data, please advise us of
any comments or questions you may have.
,
Incidentally, as discussed in a telephone conversation between Mr. R. I. Lowe and Mr. P. G. Benignus on June 1,
dS\N 20W
Addressee Monsanto Company
Date Page
June 4, 1971 3
STLCOPCB4059779
LABORATORY REPORT TO
MONSANTO COMPANY
* S$S
The following is a tabulation of the data obtained for comparative tests on a sample of a representative new oil and a sample of dodecylbenzene which we received in Belmont March 24, 1971. The samples were identified as in dicated.
Control
Tests
TOPS Limits
Representative MCS 1112 Mineral Oil
Acidity, mg. KOH/g
f
Aniline Point, C
.025 78
<0.01 44 v
<0.01 *75
Color, D-1500
0.5 1.0 <0.5
Corrosive Sulphur
' Passes
Passes
Passes
Dielectric Strength, kV D-877
26 min.
52
46
Dielectric Strength, kV D-1816
20 min.
27
26
Flash Point, C
130 min.
194
139
Fire Point, C
214 160
Interfacial Tension, dynes/cm.
40 min.
48. 3
50.2
Pour Point, F
-40 max.
Passes
Passes
Power Factor, % at 100C ' 7 :[
Power Factor, % at 25C
0.3 max. 0.05 max.
0.028 0.002
0.03 0.002
Specific Gravity 60/60 Viscosity at 100F, SUS Sludge-free Life, hours
.865 -.910 62 max. 64 min.
0.897 144 y/
140+ y
0.885
61
72
Interfacial Tension after 140 hrs . oxidation dynes / cm 15.7
17.6
REPORT BY A. R. Bompane
APPROVED BY A. M. Downey
i (C-1236)
Sheet A
DSW 201713
May 24, 1971
Doble Engineering Company Belmont, Massachusetts fj T y '5
STLCOPCB4059780
SUPPLEMENTARY DIELECTRIC BREAKDOWN MEASUREMENTS
Oxidation Time
(Hours)
0
72 140
MCS 1112 D-877 (kV) D-1816 (kV)
Representative Mineral Oil
D-877 (kV) D-1816 (kV)
52 27 46 26 34 21
31 22 32 24
(C-1236) Sheet B
* DSW 201?14
Doble Engineering Company Belmont, Massachusetts
STLCOPCB4059781
, v^DOBLE ENGINEERING CO.
rs TJrt u G
STLCOPCB4059782
DOBLE ENGINEERING CO.
STLCOPCB4059783
STLCOPCB4059784
STLCOPCB4059785
a>
0>
ri
rt4-1
Pi in
i
in
A
S
I' <
w
*
\\ '
|. w
V> I--I
VH
<
H
cWo
W Pi
Ph
w Pi
<
T3
Q 4>
>
C
rH
y
CO
u
y
fa
to CO < ui 2
CO
O
H
S.
<co
CO u
2o: 2
2
r.
t' 4 f
.8 Ih Jh
-*-> rj
W
8g
oy M* cd Ph
CQ
r-
CJ' o CO
H> -41) In O dg
td rt \u CO
y
&*
g|
H|
fa
o m n-
fa
o
in
r-
>
tt
>
X i
vO
4_> >
t0o) MTO | m fa
> P> > > >
Xi
fai
fai
fa
i
fa
m m in o m f m rf< in co
N
4-J
rH
*H
a--Ji O
y
>
rH .U
ed OB y.
Ph
a o
4r4H
O
rH
in
m v
H
u
CO a)
Q
M U
4)
to
a.
o
in cXo
n-
o CO
fa
o in n0)
6
fa
ao
o in
r-t 4-1
i-
d
*3
rH
a> X
to u
o CO
fa
o
H
u
0)
> > > >4-J
<h-H
<X
Xi
X i
fa i
Io
(\J CO
>CO <D
U rH
2 rt
U) o>
co
to
u
V
CaO . o
in cXo
d>
T3
O *H
aJ lH
d
yw
y Ur-H4
W. do
U
v
u
y o,
X ti
Oh o
co
9u
5y
od i
o rH
y rt
Oh
Ph co
y
u
Xy Ocdh
o f,
d yy
o .-
cd Oi
Ph co
d
tOd ,
8
o
r^- u
to
d
i
in
f-H
; <0
.d5
So O'
d^
fa .
co
vO O'
co
s> o Q
v
d d
i-i
fa
o
r f'-
T3
<U
>
r-l
fa
o
(U r-~-
u r-
y
p1
to <
>
X
1
vO
in
cd *-< a) d
y > > >>>
>
rH
XI XI XI XI XI
4->
OJ
o m in m m vO in m vO
d QJ
yto
h CD
Oh >
0) rH 4-)
fa (ti
bO <D
on
CO u (D
O
in cXo
DSW 201719
y
f-H
y Oh
si <d Oh O CO
o9 *y
y y
6 <od.
Ph co
SO
ma
(V) rH (L)
I 0)
U ~d
-- CO
STLCOPCB4059786
d>
4-1
ei
cr-
Qi m
t- i
in
a>
6
o co
po rt Pi W
<0 TJ
<vu Pai, o
to
Vl
ao E
5S
<ti nJ
V c/i
<0
Pi
m
r-
fNJ * Ui
V<oi
HS <
o
mt^-
4-J
<>
Xi
o oo
>
CO 00
CVOi (0
> > > > >a;
>
44
o TJ .Xrcut
44
(Q0J OiJh
Mra l h-lo->
Ji* iX Xi Xi Xi o mm o o
Vi
44
mp10
CQ
^ h o ro N
<u
Vi a>
P(0 0$
>
u
d o
aJ
o0X) )
ft
* #
Po
H
PH i
Vi
m d>
H
U m HI
(0 u d) CO a.
ion
X eg
Po
d E
*o3
! Vi
Vi 3
4->
U
0rl
--4><
w
'a
o
U
<D Vi d> Pi rP Ri
Po
to
O 'S
r 5 d)
pI
.
Pi '-i
ort
P
W
<D
m
O'
Go
er' (M
rf
nJ
*H
Pi 10 o Vi m
O3 E
o Pi o
Vi m
d>
>
X i
rl nO
H
nJ u d)
G
<D
>
rl
>>>> >
Xi Xi Xi
-X i
-X I
44
rd
4-1
o o in in in
iH CO
H
G
a>
in
<o
Vi 0)
Pi >
a>
Pi nJ oo d>
m
co u
<V CO a.
mo
X eg
d)
Vi
d) p,
rP id
cPn o
5o
to
pu
*oH Pi
nJ P W
Pa
p
o U 00
*
DSW 201720
STLCOPCB4059787
(C -1 2 3 6 )
Sheet G - C ont'd
Doble E ngineerin
June 4, 1971
GASSING TENDENCY TEST DATA
ASTM, D-2300 Procedure B
0
.
This test was performed to measure the rate at which gas is evolved or absorbed by the insulating fluid when subjected to electrical stress of sufficient intensity to cause ionization.
. Gassing Tendency
Sample Designation MCS1112 Representative Mineral Oil
Cubic Millimeters/Minute* -38 +2
Conditions of Test
Test Temperature Test Voltage Saturating .Gas Test Period
Test Method:
65 Centigrade
10-kV
Hydrogen
100 Minutes ' (between 20 and 120 minutes
elapsed from start of test -- i.e., voltage application)
Doble (Modified Thornton-Pirelli) ASTM D-2300
*
(C-1236) Sheet H
DSW 201721
Doble Engineering Company Belmont, Massachusetts
STLCOPCB4059788
DOBLE GASSING TENDENCY TEST
G. H. von Fuchs Doble Engineering Company
***
Resistance to gassing under high electrical stress has long been recognized as one of the essential properties of cable oils. Quite a number of tests have been proposed but so far none has found general acceptance, partly because of the complexity of the test apparatus and procedures involved and partly because of the lack of repeatability and re producibility of the test and the length of time required.
During 1957 ASTM became interested in developing a gassing test and the Doble Company agreed to co-operate. The apparatus proposed by Okonite (1) was a modifica tion of the Beaven-Cockburn-Thompson Cell (2) which in turn was evolved from the Pirelli cell (3). When an Okonite type apparatus was obtained, it turned out to be so complex in design that it was soon decided to make a fresh start.
One characteristic all the previous test procedures had in common was the require ment that the test oils were first to be thoroughly de-gassed and then re-saturated with the gas to be used. This required an elaborate, vacuum tight degassing and re-saturating system and resulted in expensive apparatus and cumbersome and time-consuming test procedures.
It occurred to us at the Doble Company, that if the oil to be tested could be freed of extraneous gases in the gassing cell by bubbling the test gas through it for a period of time, both the apparatus and the procedure would be simplified. Another advantage not immediately apparent, was that since there would be no requirement for creating or hold ing a vacuum, no stopcock grease need contaminate the apparatus, the oil under test be ing used as a lubricant. Among other improvements introduced by the Doble Company were a reduction in the dead gas space to the absolute minimum to speed up stripping and a specially-designed heating bath to increase precision of test results. The use of a bath also per-n: itc i testing at several temperatures.
The Doble gassing test apparatus shown in Figure 1 contains the following parts: (1) a transformer with control-box as a source of the .high-voltage a-c current, which is normally 10, 000 volts, but lower and higher voltages can also be used; (2) a thermo statically-controlled oil bath with stirrer and low-lag blade-type electric heaters, a thermometer reading to 0. 1C, and a mercury thermo-regulator connected to a super sensitive electronic relay, (3) a gassing cell and manometer assembly suspended in the bath; and (4) a pressure-regulated source of test gas which may be hydrogen, nitrogen, helium or any other reactive or inert gas or gas mixture. The operator is protected by a transparent safety shield (not shown) which is equipped with an electrical interlock. The manometer outlet is connected to a plastic tube leading out-of-doors. This is an essential safety feature when the test gas is hydrogen. The gassing cell and manometer
27AC60
LIQUID INSULATION
DSW 201722
Sec. 10-501 /<pnrrt*j<, *
STLCOPCB4059789
4 assembly shown in Figure 2 is composed of a Pyrex glass cell with the part under stress constructed of true-bore tubing. This cell has an outer ..electrode of painted-on silver with a vertical slit for observation of the oil level and a copper band for connecting to ground. The silver electrode is covered by a coating of oil-resistarrt enamel.
The principal novelty and distinguishing feature of the apparatus is the hollow high-voltage electrode which permits stripping the oil sample in the cell with the testgas at atmospheric pressure while the cell is immersed in the thermostatically con trolled oil bath maintained at test temperature. Thus the need for vacuum tight, com plicated auxiliary equipment normally used for de-gassing the oil sample is eliminated; and this operation, as well as saturation of the oil sample with the test gas, is accom plished in a greatly simplified single step. Therefore considerable time is saved and the two operations can be completed in 20 minutes instead of the several hours which were required previously.
The manometer is connected to the side-arm of the gassing cell by a standard taper ground-glass joint and held in place by spring clamps. The manometer has an etched-on millimeter scale and is so constructed that the manometer-liquid (di-butylphthalate) can not contaminate the oil in the test cell.
A schematic drawing of the cell and manometer assembly is shown in Figure 3.
During.part of the stripping operation, the test gas, which enters the cell at the lpwer end of the inner electrode, passes through the test oil and through the manometer liquid. This is intended to remove all extraneous gas from bulb A and also explains the need for bulb C. Bulb A serves as a reservoir for the manometer liquid when hydrogen absorb ing oils are tested, while bulb B permits the use of the full length of the graduated manometer arm when gas is evolved. The stopcock is used both to by-pass the mano meter during the last ten minutes of the stripping period and to equalize the manometer liquid levels after stopping the gas flow.
' Detailed dimensions of both the glass cell and the inner (high tension) electrode are shown in Figure 4. This electrode is made of 10 mm o. d. centerless ground stain less steel seamless tubing and contains a brazed in stainless steel capillary tube as gas passage. The electrode is supported and centered in the glass cell by a precision machined, recessed 24/40 tapered Teflon plug. A 1/8" needle valve with gas inlet is on top of the electrode.
The test procedure is as follows: 4. 8 ml of the oil to be tested is added to the cell and the electrode inserted using a few drops of the test oil as lubricant and sealant. The cell and manometer assembly is suspended in the oil bath previously heated to the test temperature (95C). The gas inlet and outlet connections are attached and the test gas passed through the oil and through the manometer liquid as a steady stream (about three bubbles/second) for ten minutes. Following this, the bypass^ stopcock of the manometer ^s opened and gas stripping and saturation of the oil sample is continued for another ten minutes. Then the gas supply and the electrode inlet valve are turned off and the by^pass stopcock is closed, the high-voltage and ground leads are connected to the cell and the safety shield is lowered. Before turning on the current, the manometer level is read and the readings repeated at one minute intervals until two successive readings give the same value, which is recorded. Then the current is turned on by gradually increasing the voltage until the test voltage (10, 000 volts) is reached, this voltage being maintained throughout the test. The ASTM testing time is normally 100 minutes and is preceded by a 20-minute run-in period. The alleged purpose of this run-in period is to arrive at a
Sec. 10-502
LIQUID INSULATION )
DSW 201723 27AC60
STLCOPCB4059790
relatively steady reaction rate. Thus in our' routine gassing test, manometer readings
are taken 20 minutes and 120 minutes after the current is turned on and the difference
is recorded in millimeters per 100 minutes.
'
.
9.
If a decrease in gas volume is observed such as with oils absorbing hydrogen in
the test, the results are marked with a negative sign.
;
The test results are expressed in cubic millimeters per minute which is arrived at by multiplying the millimeter change in liquid level by the manometer constant and dividing the result by the test time in minutes. In our case the manometer constant is 25 cubic millimeters per millimeter change in manometer level.
Average gassing rate = ^2 ~ ^1 x 100
= ^2 ~ ^1 = cubic mm/minute 4
V2 is the reading of the manometer at 120 minutes.
Vj is the reading of the manometer at 20 minutes.
Gassing test values for fourteen commercial oils with hydrogen and with nitrogen -as test gases are shown in Table I, which also contains the specific optical dispersion (4) values of these oils.
While hydrogen is the generally accepted test gas, we at the Doble Company be lieve that the use of nitrogen is more realistic. Hydrogen is hardly ever encountered in normal service, while nitrogen is the principal constituent of cavities believed to be the initiators of the gassing reaction. Nitrogen is also commonly used as an inert gas blanket in electrical equipment.
In the gassing reaction, hydrogen is split off from hydrocarbons 'under electrical stress. Saturated hydrocarbons which are the principal constituents of over-refined oils lose hydrogen much more readily than do the so-called "unsaturates" or aromatics. The chemical action of electrical stress is comparable to that of thermal stress which in the petroleum industry is referred to as "cracking". Due to their unsaturated nature, the thermally and electrically more stable aromatics are capable of chemically absorb ing the hydrogen atoms split-off from the saturated hydrocarbons before these atoms have a chance to combine to molecular hydrogen. Since gassing is largely a vapor phase reaction, oils containing the right kind of (light) aromatics will resist gassing.
Specific, dispersion is a measure of total aromaticity and thus does not always
correlate with gassing tendency. Examples of such inversions are Samples 1 and 2 and Samples 13 and 14. In the case of Samples 1 and 13, it is assumed that the oils have less of the light aromatics than their counterparts. Samples 2 and 14. According to its specific dispersion value (100. 4) Sample 2 is very low in aromatic content: it is prac tically a white oil. Thus it is interesting to note that Sample 1 with a substantially higher aromatic content gases more, when either hydrogen, or nitrogen is used as i test gas.
While the right kind of aromatics are capable of absorbing activated hydrogen atoms and thus make an oil an absorber when using hydrogen as a test gas, certain highly re active olefinic compounds absorb hydrogen even more readily. An experimental cable oil of low aromaticity but containing such a reactive, presumably olefinic, compound as an additive, was capable of absorbing more hydrogen than was a commercial cable oil of
27AC60
LIQUID INSULATION
DSW 201724 Sec. 10-503
STLCOPCB4059791
much higher aromatic content. However, when the same two oils were tested with nitro
gen, as shown in Table II, the situation was reversed, Sample 1 formed more gas than
Sample 2.
' ,.
These results and similar confirmatory work with argon and helium as test gases convinced us at the Doble Company, that gassing tests with hydrogen are not always re liable because the test can be passed when additives are present which are capable of absorbing activated hydrogen when it is used as a test gas and is thus present in excess; but these same additives do not readily absorb the hydrogen atoms split off from saturated hydrocarbons under electrical stress.
Nitrogen activated in an electric field is not completely inert. In the gassing test with a highly aromatic oil, we were able to show that nitrogen is absorbed giving nega tive test values. When the same oil was tested with truly inert gases such as argon or helium, the manometer level remained constant even when the test was extended over 8 hours (500 minutes) showing that no hydrogen was liberated.
These findings further justify the Doble Company's preference for electrical oils
which contain light aromatic hydrocarbons, since the proper type of aromatics will ab
sorb nitrogen as well as hydrogen, and such oils will tend to decrease or will even
eliminate voids in cables by absorption and thus further decrease the chances for gas
formation under electrical stress.
*
In conclusion, the Doble Gassing Tendency test presented here shows considerable promise. The design of the gassing cell is simple and the test is easily repeatable and reproducible.
As evidenced at recent ASTM meetings, there is considerable interest in a prac
tical gassing test for cable oils. While some of the transformer people are not as yet
concerned with the need for such a test, improvements in transformer design will bring
about this requirement for transformer oils as well. In the meanwhile, the use of non
gassing transformer oils should be regarded as additional insurance for trouble-free
service performance. This is expressed by the Doble Company's inclusion of the gassing
tendency test in their oil specifications.
_
' . REFERENCES -
(1) S. C. Bartlett and R. B. Blodgett; "A Correlation Between Gas Evolution
in Electrically Stressed Oils and Other Estimates of Oil Aromatic Con
tent" Paper presented at the Twenty-Eighth Annual Meeting, N. R. C. Con
ference on Electrical Insulation, October 26-28, 1959, Pocono Manor Inn,
Pocono Manor, Pa.
*
(2) G. H. Beaven, J. A. Cockburn and C. N. Thompson; "The Evaluation of* Gassing Tendencies of Insulating Oils: Apparatus, Procedure and Effect of Experimental Variables", Journal of the Institute of Petroleum, Vol. 35, No. 311, p. 735-754, November 1949.
(3) G. Palandri and A. Pelagatti; "Insulating Oils for Electric Cables" paper presented at Convention of the Associazione Elettrotecnica Italiana in Bellagio, Italy, October 1954.
SVV 201725
Sec. 10-504
LIQUID INSULATION
27AC60
STLCOPCB4059792
(4) G. H. von Fuchs and A. P. Anderson; "Effect of Solvent Extraction on
Aromaticity of Lubricating Oils" Industrial and Engineering Chemistry
Vol. 29, pages 319-324. (March 1937)
.
TABLE I
Doble Gassing Tendency Test -- (120-20 min) at 95 C
Sample Cubic mm /min. Aromatics
No. H2
n2
Spec. Disp. *
1 +34. 5 59.9 2 +32. 6 44. 8 3 +9.2 33. 4 4 +8. 0 29.8 5 +8. 0 34. 9 6 -0.4 24.6 7 -1.5 28. 6
105. 8 100. 4 107. 0 109.0 111. 1 113. 2 111. 2
Sample Cubic mm/min. Aromatics
No. h2
n2
Spec. Disp. *
8 -7.8 18.4 9 -8. 6 21. 0 10 -11.2 23. 6 11 -11.2 20. 0 12 -11. 5 18. 8 13 -18. 6 21. 1 14 -27. 0 14. 8
114. 4 115.4 114. 4 114. 9 114. 4 119.7 117. 0
* Specific Dispersion by Dispersometer
TABLE II
Doble Gassing Tendency
Sample No. h2
N2
1 -20. 1 +27. 3 ' 2 -8. 9 + 17. 3
Aromatics Specific Dispersion*
100. 5 115. 5
* By Disper someter
-
Note: Sample 1 contains a hydrogen absorbing additive
27AC60
ft LIQUID INSULATION
Sec. 10-505
STLCOPCB4059793
j-n! f^rs
L Sr-
ns
Sec 10-506
! 1^'
Doble Gassing Test Apparatus
FIGURE 1
LIQUID INSULATION t
27AC60
STLCOPCB4059794
Schematic Diagram of Cell and Manometer Assembly
FIGURE 3
DSW 201728
Sec. 10-508
LIQUID INSULATION
27AC60
STLCOPCB4059795
t
27AC60
1
-4
i "i
* i
3
.i
)
A
* i *
Gassing Cell and Manometer Assembly
figtoe 2
LIQUID INSULATION
Sec. 10-507
STLCOPCB4059796
C aM Zhtb
Detailed Dimensions of the Glass Cell and the Inner (High-Tension) Electrode
t
FIGURE 4
' DSW 201730
27AC60
LIQUID INSULATION
Sec. 10-509
STLCOPCB4059797