Recent Developments in The Interpretation: of Dissolved Gas Analysis in Transformers
Recent Developments in The Interpretation: of Dissolved Gas Analysis in Transformers
TECHNICAL BROCHURE
296
GT D1.01/
Recent developments in the interpretation A2.11
of dissolved gas analysis in
transformers
296
TECHNICAL BROCHURE
Table 1: Ranges of 90 % typical concentration values for It thus appears that stray gassing in gen-
GT D1.01/
power transformers, in ppm
eral will not significantly interfere with DGA A2.11
C2H2 H2 CH4 C2H4 C2H6 CO CO2 diagnoses in service, except when a strongly
All 50- stray gassing oil is used (such as new Voltesso
transformers 150 30- 60- 20- 400- 3800-
35), or when transformers are operated well
130 280 90 600 14000
No OLTC 2-20 above nominal load (IEC hottest-spot tem-
Communicating 60-280 perature at nominal load, for example, is
OLTC 98 °C). Heat run tests in general will not be
affected by the stray gassing of oil.
Table 2: Ranges of 90 % typical rates of gas increase for power
transformers, in ppm/year
DGA users are recommended to ver-
C2H2 H2 CH4 C2H4 C2H6 CO CO2
All 35-132 10-120 32-146 5-90 260- 1700- ify the stray gassing tendency of their oils
transformers 1060 10,000 at the hottest spot temperature of their
No OLTC 0-4 transformers, which depends on load con-
Communicating 21-37 ditions and ambient temperature. Tests are
OLTC described in the report for that purpose,
and methods to calculate the contribution
of stray gassing to gas formation in service. Criteria are
Individual networks are strongly recommended to also indicated to identify stray gassing in service and
calculate their own typical values. Methods are provided distinguish it from other similar types of gas formation
in the Brochure for that purpose. DGA users are in service (catalytic reactions on metals, partial dis-
reminded that typical values are not limit values requir- charges, low-temperature thermal faults).
ing immediate action on the transformers but only more
frequent monitoring by DGA. Gas formation from paper
Gas formation from paper at different temperatures
Typical values are not significantly affected by the oil (125 to 250 °C) has been measured and reported in the
volume or size of transformers (except for very small report, allowing to calculate the amount of paper
ones). They are higher in the early years of the equip- involved in a fault when its temperature can be estimated.
ment, suggesting that some unstable chemical bonds of
oil and paper are initially broken then stabilize. Typical
values are higher in shell-type transformers and shunt Gas formation from partial discharges
reactors, possibly because these usually operate at higher It has been shown that DGA is not as sensitive as elec-
temperatures. They are lower in instrument than in tric or acoustic tests to detect very low levels of PDs. How-
power transformers. ever, DGA is more useful to detect the PD levels which
just become harmful to the equipment.
Pre-failure values in service
Gas concentration values observed just before fail- Gases trapped in paper insulation
ure of transformers in service have been surveyed and These gases are very difficult to remove completely,
found to be surprisingly similar on different networks, requiring several years in service after re-commissioning
suggesting that failures occur when a critical amount or months by vacuum treatment. It is therefore preferred
of insulation has been destroyed by a fault. Ranges of pre- in the industry to use the patterns of gases released in the
failure values are indicated in the report. oil to detect newly occurring faults by subtraction. ■
Thermal stray gassing of oil Table 3 : Calculated contribution of stray gassing to gas levels in
This refers to the unexpected formation transformers after 1 year in service, in ppm of H2