C57 104
C57 104
104-1978
7
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
8
ANSI/IEEE
OF GENERATED GASES IN OIL-IMMERSEDTRANSFORMERS C57.104-1978
in nitrogen can, therefore, be employed to indicate gas content is probably above the upper explo-
the degree of detector element fouling.
sive limit. This can be verified by additional
2.6 Calibration. dilution of the sample gas with air to bring the
2.6.1 Prepare the instrument for operation combustible gas content within the range
and make zero balance and voltage adjustments of the instrument.
in accordance with the instructions of the instru- 2.7.6 Recheck the instrument’s zero balance
ment manufacturer and the procedure, 2.7. and voltage adjustments.
2.6.2 Mix the standard reference gas with air 2.7.7 Report reading in percent to the
and introduce it into the meter in exactly the nearest small-scale division.
same manner as used for the sample gas from 2.7.8 After use, purge the instrument to
the transformer. remove sample gas in order to minimize cor-
2.6.3 Turn the calibration adjustment so the rosion and eliminate contamination.
meter indicates the known methane content of
the reference standard when the air and
reference gas mixture is being tested. Purge the 3. The Use of Fixed Instruments for the
meter with air to remove traces of reference gas
and proceed as in 2.7. Detection and Determination of the
Quantity of Combustible Gases
2.7 Procedure.
2.7.1 Prepare the instrument for operation Present in Gas-Blanketed
in accordance with the instrument manu- Equipment
facturer’s instructions. This should normally
include a warmup period to allow the battery The reliability of sealed transformers with a
voltage to stabilize, a voltage adjustment, and gas space over the oil can be enhanced by
an instrument balance or zero adjustment. monitoring the gas space using self-contained,
2.7.2 Calibrate the instrument against the fixed mounted instruments. These instruments
standard methane mixture following the detect and determine the quantity of com-
procedure in 2.6.The frequency of calibration bustible gases as well as sound an alarm when
required for individual instruments can be the combustible gases exceed a predetermined
determined by experience. Low readings level. Optional recorders can also be used to
because of contamination on the platinum provide a daily record of the combustible gases
elements can only be detected by a check with present.
a gas of known composition. In any case of 3.1 Scope.
doubt, the instrument should be recalibrated. 3.1.1 This method covers the detection and
2.7.3 If the instrument has a range switch determination of combustible gases in the gas
or more than one dilution orifice, set for space above the oil in transformers using fixed
maximum range or dilution to avoid damage to instruments and sounding an alarm when the
the instrument. Purge the instrument so it is combustible gases exceed a predetermined
free of gas from previous tests. level that is considered to be indicative of
2.7.4 Introduce the mixture of air and pending trouble.
sample in accordance with instructions of the 3.1.2 This method affords a quantitative
instrument manufacturer. The flow rate measurement of the total combustible gases
should be moderate and reproducible. The present in the transformer gases. If the amount
pressure of gas in the oxidation chamber of the individual gas components is desired, a
should be atmospheric and the temperature laboratory analytical method using a gas
should be above the dew point of the gas being chromatograph or mass spectrometer should
tested to prevent condensation of moisture on be used.
the oxidizing elements.
2.7.5 Read the percent combustible gas 3.2 Summary of Method.
indicated on the meter. If the reading is low, 3.2.1 At least two somewhat related methods
change to a lower range or change the orifice are employed.
to one admitting less dilution air. If the instru- 3.2.1.1 The first type of gas monitor
ment goes momentarily off scale on the high continually compares the thermal conductivity
side and then returns to zero, the combustible of the transformer gas with that of pure
9
ANSI/IEEE
(257.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
nitrogen and is suitable for any transformer of of its formation over the oil resulting from 1
the closed type with a gas space above the incipient internal fault in transformers. If the
transformer oil. level of combustible gases reaches a preset limit
3.2.1.1.1 The transformer gas is con- an alarm is actuated, this warning of pending
tinually circulated through one section of a trouble.
Wheatstone bridge and returned to the trans-
former. The other section of the bridge contains 3.4 Apparatus.
pure nitrogen and is balanced against the 3.4.1 Fixed Combustible gas apparatus’
transformer gas. consisting generally of the following com-
3.2.1.1.2 When combustible gases are ponents.
produced in the transformer, they mix with the (1) A Wheatstone bridge circuit capable of
transformer gas and increase the thermal accurately determining changes in thermal
conductivity of gases, with controls for bal-
conductivity of the transformer gas. The
ancing and voltage adjustment.
increase in the thermal conductivity of the
transformer gas unbalances the Wheatstone (2) A milliampere meter calibrated to read
directly in percent total combustible gases.
bridge and is proportional to combustible
3.4.2 The standard reference gas shall be
gases as indicated on a meter.
made in accordance with manufacturers’ rec-
3.2.1.2 The second type of gas monitor
ommendations.
continuously samples the transformer gas at
3.4.3 Optional Recorder. Provision is made
fixed intervals and bums any combustible
on the fixed combustible gas apparatus to
gases present to provide a measure of the
install a recorder to give a permanent daily
combustible gases. This type of monitor is
record of changes in combustible gas
used only on transformers with a positive
concentration.
pressure of nitrogen over the oil.
3.2.1.2.1 At a fixed interval (usually 3.5 Safety Precautions. Gases generated in
24 h) a sample of the transformer gas is pumped transformers can be explosive. Strict precau-
from the unit, mixed with air, and passed tions should be observed when sampling the
_-
over a heated platinum sensor of a Wheatstone gases from transformers.
bridge. Any combustible gas in the sample is
burned. This raises the temperature of the sensor 3.6 Calibration. The instruments shall be
and unbalances the bridge which was balanced calibrated in accordance with the manufac-
against a second platinum sensor in air. The turers’ instruction guide.
degree of unbalance is proportional to the 3.7 Maintenance.
amount of combustible gases present in the 3.7.1 At the time of installation and each
transformer gas as indicated on a meter. year thereafter, the equipment should be
3.2.1.3 A relay is operated in conjunction standardized to be sure the monitor is operat-
with the percent gas meter so that when the ing properly.
combustible gases exceed a preset value (usually 3.7.2 For restandardization and repairs, the
1percent) the relay sounds an alarm. operator should follow the manufacturers’
3.2.1.3.1 The alarm relay can be reset at instruction guide.
a higher combustible gas level once the initial
1 percent level has been reached. The time to
reach this new level will give a measure of the 4. Procedures for Obtaining Samples of
rate of gas formation. Gas and Oil from the Transformer for
3.2.1.32 All units that have sounded an Laboratory Analysis
alarm should be checked with a portable gas
meter to verify the combustible gas level, or 4.1 Gas Samples for Laboratory Analyds.
the gas should be sampled for complete analy- 4.1.1 Gas Blanket. It has become common
sis by a gas chromatograph or mass spec- practice to extract samples of gas from the
trometer.
3.3 Significance.
3.3.1 The combustible gas monitor provides 2 A commercial instrument found to be satisfactory
a means for the continuous detection of com- is TCG Fault-Gas Monitor, available from McGraw-
bustible gas and the determination of the rate Edison CO, Power Systems Division, Canonsburg, PA.
lo
ANSIIIEEE
OF GENERATED GASES IN OILIMMERSED TRANSFORMERS C57.104-1978
TUBING
ADAPTER
HOKE OR
w CAP WITH
SEALING GASKET
SIMILAR
VALVE
AND FITTINGS
-STEEL
CYLINDER
CAPACITY CAN CAPACITY
UP TO 1000 rnL OVER 1000mL
Fig 1
Metal Sample Containers
gas blanket above the oil in sealed transformers, Sample size requirements are a function of
for submission to the laboratory for quantita- the sensitivity of the apparatus in use and the
- tive analysis. This may be conveniently amount of dissolved gas present. Samples as
performed in accordance with ASTM D2759, small as 1 ml or as large as 1000 ml have
Standard Method for Sampling Gas from a proven to be satisfactory with current
Transformer under Positive Pressure, being technology.
careful to obtain a representative sample. 4.1.4 Transport and Direct Sampling Con-
4.1.2 Gas Collection Device. Transformers tainers. It is essential that the oil sample be
not having a gas space may be equipped with a representative of oil within the equipment.
gas collector relay which is connected to the Following are recommendations for the accep-
highest point of the transformer in order to tance criteria for transport and direct sampling
collect the gases which may form in the unit. containers.
A gas sample may be taken from such a device 4.1.4.1 Transport Containers. Transport
by connecting a gas-tight (or precision ground) containers should meet the leak criterion
glass syringe through a short length of plastic in 4.1.4.2. They should be suitable for ship-
tubing and a three-way stopcock t o the device. ment by mail and by any freight carrier. It
To do so, flush the tubing, stopcock, and cannot be overemphasized that the results of
syringe by valving the gas from the relay to the the analysis by this method may dictate
syringe. (See ASTM D923, Tentative Method judgments on the serviceability of power
for Sampling Gas from a Transformer by equipment; therefore, all precautions should be
Syringe.) taken to ensure the safe delivery of good
4.1.3 Gas Dissolved in Oil. Gas decomposi- uncontaminated samples.
tion products of electric insulating oil and Several types of containers are presently
solids are soluble in oil to varying degrees used. They are listed below.
depending on composition.
In the case of oil-filled (diaphragm or con- (a) calibrated stainless steel cylinders (Fig 1)
servator) transformers, until sufficient gas has (b) flexible metal cans (Fig 1 - one sug-
accumulated to operate a gas detector relay, gested type)
(c) syringes (see ASTM D923)
there is no gas in the transformer tank above
(d) glass bottles.
the oil; hence, the only gas available for analysis
is that dissolved in the oil. The optimum method of filling the can and
11
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
sampling the contents of the can are still under in Fig 1. Syringes must also conform to the
development. leak criterion in 4.1.4.2 and be used in accor-
4.1.4.2 Leak Acceptance Criterion for dance with ASTM D923.
Transport Containers. The introduction or loss 4.1.4.4 Oil Sampling Using Metal Bottles -
of gases to or from the sample during sampling, Oil Sampling Procedure. Tubing adaptors are
transport to the analytical laboratory, and attached to each end of the threaded ends of
transfer to the extraction process should not the shutoff valves of the stainless steel sampling
exceed 0.35 percent of the hydrogen or helium bottle. Two lengths of flexible polyvinyl
content of a standard sample per day -in lieu chloride tubing are attached to the tubing
of other procedures meeting this criterion, adaptors. Before attaching the polyvinyl
ASTM D923 is recommended, but subject to chloride tubing to the main sample valve,
the above leak criterion. flush the valve into a scrap-oil container. One
4.1.4.3 Direct Sampling or Transfer Con- end of the tubing is attached to the oil sampling
tainers. Glass syringes are the direct sampling valve of the transformer and the open end of
containers of choice where the oil is taken the other piece of tubing is placed in a scrap-
directly from the equipment and introduced oil container. Both valves of the sample bottle
into the extraction device. They are also are opened and then the main valve of the
recommended for transferring oil or gas samples transformer is opened. The sample bottle
from transport containers (4.1.4.1) as shown must be held in the vertical position (Fig 2).
Fig 2
Oil Sampling with a Metal Bottle
12
ANSI/IEEE
OF GENERATED GASES IN OILIMMERSED TRANSFORMERS (357.104-1978
- The oil will gradually fill the oil container from way high-vacuum stopcock lubricated with the
the bottom and overflow through the top minimum amount of high-vacuum silicone
into the scrap-oil container. Continue over- grease.
flowing the oil until no more bubbles can be A degassing flask with a fritted glass gas
seen through the plastic tubing and until two dispersion inlet tube, of sufficient volume to
quarts of oil have been collected in the scrap- contain up to 50 ml of oil below the inlet
oil container. Close the top valve of the sample tube, capable of being evacuated, containing
bottle and allow oil pressure to build up, then a ptfe coated magnetic spin bar, and mounted
close the bottom valve of the sample bottle on a magnetic stirrer.
and finally the main transformer valve. A means of measuring absolute pressure
within the apparatus.
NOTE: Oils having a high degree of gas saturation A vacuum pumping system capable of evacu-
may account for bubbles if there is a significant change
in the temperature in the sample in or during transit. ating the glassware to an absolute pressure of
(See 5.1.2.4.4.) 1 X 10-3 torr or lower.
Vacuum glassware sufficiently large com-
pared to the volume of the oil sample, so that
virtually complete degassing is obtained and
5. LaboratoryMethods for Analyzing the the volumetric collection ratio is as large as
Gas Blanket and the Gases Extracted possible. A 500 ml gas collecting flask has been
from the Oil found suitable.
High-vacuum valves or stopcocks employing
5.1 Gas Extraction and Analysis. the minimum necessary amounts of high-
5.1.1 Gas Blanket Sample. The sample of vacuum stopcock grease are used throughout
the gas blanket may be submitted directly to the apparatus.
analysis as described in 5.2.3 except that the A gas collection tube calibrated in 0.01 ml
results will be reported as percentages. divisions capable of containing up to 5 ml of
- 5.1.2 Gas Extraction from Oil Sample by gas, terminated with a silicone rubber retaining
Expansion in an Evacuated Flask. septum. A suitable arrangement is shown in
5.1.2.1 Scope. This method covers the Fig 5.
extraction and measurement of gases dissolved A reservoir of mercury3 sufficient to fill the
in electrical insulating oil having a viscosity of collection flask and collection tube.
20 centistokes (100 Saybolt Universal Seconds) 5.1.2.4.2 Determination of apparatus
or less at 4OoC (104'F) and their containment collection ratio. The apparatus shall be care-
for quantitative analysis. Any noncondensable fully checked for vacuum tightness of all
gas may be determined. joints and stopcocks.
5.1.2.2 Summary of Method. Dissolved The volume of the total extraction space
gases are extracted from a sample of oil by V T , and the volume of the collection space
introduction of the oil into a pre-evacuated V c , shall be measured, and the ratio
known volume. The evolved gases are com- VC
where V, is the volume of
pressed to atmospheric pressure and the total (VT -k vcyl)-v,
volume measured. oil sample, VT is the total internal volume of
5.1.2.3 Description of Terms. Gas content extraction space, and Vcyr is the volume of
of oil by volume is the total volume of gases, calibrated stainless steel cylinder (if a syringe
corrected to 760 torr and zero degrees Celsius, is used this number is zero) shall be calculated
contained in a given volume of oil, expressed as the volumetric ratio.
as a percentage. The degassing efficiencies for each individual
5.1 2.4 Appamtus. component gas shall be calculated using the
5.1.2.4.1 Method A: A method suitable formula
1
for determining the total amount of dissolved
gas and the amounts of the individual dissolved
gases using samples of 1 ml to 50 ml in size.
- (See Figs 3 and 4). Flexible polyvinylchloride
tubing terminated with a Luer fitting to accept ~~
a glass syringe and leading t o a solid plug, three- Suitable handling precautions shall be observed.
13
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
STAINLESS
STEEL
CYL IN DER
PROCEDURE
5. I .2.4.4
EXTRACT GAS SAMPLE
WITH SYRINGE
I
Fig 3
Apparatus for Extraction of Gas from Insulating Oil
Method A Type I
14
ANSI/IEEE
OF GENERATED GASES IN OILIMMERSED TRANSFORMERS C57.104-1978
15
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
.
,x.,
.U
i \
STEEL CYLN
STAINLESSI- i
PROCEDURE
5.1.2.4.4
Fig 4
Apparatus for Extraction of Gas from Insulating Oil
Method A Type I1
COLLECTION TUBE
mine total dissolved gas. That value must be
determined by an independent method. The
limitation of this method is being investigated
(see Fig 6).
Fig 5
Retaining Rubber Septum for Gas
Collection Tube
16
ANSI/IEEE
OF GENERATED GASES IN OIL-IMMERSED TRANSFORMERS C57.104-1978
GAS COLLECTION
TUBE
TO SAMPLING
POSITION I
VEgsEL VESSEL
C" "A"
500 CC 500 CC
POSITION 2
I I
TO VACUUM
Fig 6
Apparatus for Extraction of Gas from Insulating Oil
Method B
5.1.2.4.6 Procedure. Set stopcock A in and B.
position 3, and open stopcocks B, C, D, E, F, Partially open stopcock B to allow the oil
and G. Close stopcock H. to flow into vessel A. 3
Connect sampling bottle to stopcock A. When vessel A is approximately 4 full close
Evacuate the system to an absolute pressure stopcock A by turning it to a position half
of 10-2 torr or lower. way between position 1and position 2.
Close stopcocks B and C. Close stopcocks D and E. Turn stopcock A
-. Turn stopcock A to position 2, open the to position 1.
valves on the sampling bottle and allow the Open stopcock C.
oil to fill the tubing between stopcocks A When vessel B is full of oil,close stopcock F.
17
~
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
18
ANSI/IEEE
OF GENERATED GASES IN OIGIMMERSED TRANSFORMERS C57.104-1978
19
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
20
ANSI/IEEE
OF GENERATED GASES IN OIL-IMMERSED TRANSFORMERS C57.104-1978
0.5 2.1
to to
G0.49 2.0 5.0 >5.0
The initial level of total combustible gas and the rate of increase of generation are calculated. The appropriate
testing and surveillance procedures recommended in Table 3 are:
Procedure A: Analyze weekly. If rate remains the same or decreased, analyze monthly and then bi-monthly. If
gassing then stops check 4-6 months.
Procedure B: Analyze every 3 days. If rate is constant, check for load dependency. If load dependent, reduce
load if possible and monitor weekly.
Procedure C: Analyze daily. Check load dependence, audible internal corona. If rate drops use Procedure B and
A.
Procedure D: Take gas space sample for analysis.
Procedure E: Take oil sample for analysis.
venient system outages for replacing faulty evaluate the results of the analyses per 6.3.2.
equipment necessitates early detection and (2) Determine the total combustible gas
immediate information on the generation rate generation rate by periodic sampling or by
or sensitivity of the generation rate to changes recording results from continuous monitor.
in load. When combustible gases are first (3) Determine equipment serviceability and
detected in the gas space, relay, or dissolved the appropriate action by utilizing the informa-
in the oil it is recommended that the equip- tion on TCG levels from gas space samples only
ment manufacturer or service representative be and generation rates as described in Table 3.
notified immediately. Then depending on the
type of gas detection used, TCG or dissolved 6.3.2 Determination of the Operating Condi-
gas, the following procedures recommended in tion of Transformers by Analysis of the
this guide can be utilized t o assess the operat- Separate Gases Extracted from the Oil or
ing condition of the equipment. Removed from Relay or Gas Space. Again it
6.3.1 Determination o f the Operating Condi- is recommended that the equipment manu-
tion o f Transformers Using Total Combustible facturer or his service representative should be
Gas (TCG) Analysis (Sections 2 and 3). When notified when fault gases are first detected.
either portable devices or fixed devices indi- While a number of methods are known to be
cate combustible gas: in use for the interpretation of the gas analysis,
- (1) Remove a sample of the gas from the two of the most commonly used methods are
gas space (or relay) or obtain an oil sample, or qualitative interpretation (6.3.2.1) and ratio
both. Analyze per Section 5 of this guide and and rate analysis interpretation (6.3.2.2).
21
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
These methods used in conjunction with TCG testing procedures will provide new informa-
methods (6.3.1) provide the equipment tion to permit better maintenance techniques
operator with valuable insight into the loca- for transformers.
tions and the types of faults generating the 6.3.23 Ratio and Rate Analysis Znterpre-
gases. They also provide further guidance tation. Two most useful ratio methods
concerning the gravity of the fault and a better attributed to Domenberg (6.3.2.2.1) and
estimate of the progress of the fault and the Rogers (6.3.2.2.2) utilize the ratios of key
relative urgency of a planned outage. Accurate component gases obtained from gas space,
fault type analysis by these procedures relay or extracted from the oil to diagnose the
expedites repair. fault type. The literature (7.2)contains con-
6.3.2.1 Qualitative Interpretation. While siderable documentation of actual cases which
some investigators attach more or less impor- provide the basis for the assignment of fault
tance to the detection of certain gases, a types to these ratios.
generally accepted list of gases and the condi- 6.3.2.2.1 Dornenberg mtio method (re-
tion or type of fault generally associated with vised 1972). The gases considered sufficient
them can be found in Table 4. to describe types of faults are: H 2 , CH4, C2H4,
It is recognized that there are many variables C2H6, CO, and C O 2 . Units of moles/liter or
that prevent clear-cut interpretation of faults ppm v/v can be used for calculating the ratio.
found in transformers and the steps that must Three techniques are applied: key components,
be taken to eliminate the gassing problem. norms, and ratios.
However, in many cases, the technique of Key Components:
continuous monitoring and analyses of dis- (a) thermal decomposition (hot spots) :
solved and free gases generated in transformers mainly C2H4, CH4; less C2H6, H 2 ;
has become a valuable tool in the diagnosis of sometimes C2H2
the transformer fault and how best t o correct (b) electrical discharges (except corona)
it. There is little doubt that improvements in mainly H 2 , CH4 (H CH,); C2 gases in
Table 4
Types ofProbable Faults
22
ANSI/IEEE
OF GENERATED GASES IN OIL-IMMERSED TRANSFORMERS C57.104-1978
Table 5
Dissolved Gases (GasExtracted from Oil)
Domenberg Method
-
CZHZ -
CZH6 -
CZHZ
CZH4 CZH2 cb
Thermal
decomposition
(hot spots)
1 1
>l.o
<0.75
1 1
>0.4 <0.3
Thermal
decomposition
(hot spots) >0.1 <l.O x.2 <O.l
Electrical Electrical
discharges discharges >O.Ol
(except corona) <0.1 >l.o <0.2 X.1
Corona
significant
Corona I
I
<O.OI I
I
not
significant
1 1
I
>O.Z
II
<O.I
I I I I
Table 7
Suggested Diagnosis from Gas Ratios-Rogers Ratio Method
7
-
CH, CZ H4 cz Hz
H2 cH4 cz Hci c2 H4 Suggested Diagnosis
m.1
<1
(0.1
.o I <1 .o
<1
<1.o
.o
I <0.5
I Normal
<O.l
>0.1
II <l.O
<l.O
23.0
I 2:;
23.0
or 23.0
I Partial discharge - Corona
with tracking
Continuous discharge
<l.O
I 7
>0.1
'
<l.O
2::; or 23.0
I
20.5 or 23.0
<3.0
I
Arc -With power follow
through
I
I I
Arc - No power follow
<1.o
I <1.o 20.5
<3.0 through
2::: or 23.0
I I I I Slight overheating - t o
15OoC
213; or >3 .O
I >1 .o I <l.O I <0.5
I Overheating 150-200°C
>O.l
<1.0 I 21.o
<3 .O I <Oa5 I General conductor over-
heating
23
ANSIIIEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
24
ANSI/IEEE
OF GENERATED GASES IN OIL-IMMERSED TRANSFORMERS C57.104-1978
25
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
26
ANSI/IEEE
OF GENERATED GASES IN OILIMMERSED TRANSFORMERS C57.104-1978
27
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
28
ANSI/IEEE
OF GENERATED GASES IN OIL-IMMERSEDTRANSFORMERS C57.104-1978
29
ANSI/IEEE
C57.104-1978 IEEE GUIDE FOR THE DETECTION AND DETERMINATION
30
ANSIIIEEE
OF GENERATED GASES IN OIL-IMMERSED TRANSFORMERS C57.104-1978
31