in this issue - Electricity Today Magazine

Transformer Study
Continued from Page 52
The calculated evacuated volume is
larger for small and severe short-circuits.
To explain this result, it is important
to remember that:
• The pressure set point is the same
for all calculations and therefore the gas
quantity produced before pressure
reaches 0.8 bars is the same for all
examples.
• The only difference at the instant
of opening is the pressure gradient.
• As explained in section 6, “Analysis of the DEPRES-
SURIZATION CHAMBER rupture disk efficiency”, the disk
opening time varies according to the pressure slope.
• For the 236kA short-circuit current, the rupture disk
opening time is 0.6 ms whereas it is 1.8 ms for the 35 kA fault.
• The opening speed is crucial to the depressurization
process. The shortest opening time provokes the fastest pressure
drop.
Hence, a quick depressurization provoked by a fast pressure
rise, which gives a rapid rupture disk opening time will
only require evacuating a small volume of oil.
Figure 18: Depressurization chamber efficiency analysis for a
236kA short-circuit, pressure in bar and psi
9. DEPRESSURIZATION CHAMBER
CALCULATION
The length of the depressurization chamber is designed
according to the amount of oil to be expelled during the depressurization
process, calculated in section 8.
However, an important fact that does not appear in section
8, “Evacuated Volume Calculation” and the figures presented in
section 7, “Depressurization Calculation” is that when the rupture
disk has operated, the generator is still feeding the fault in
any case, and there might be a time lag before the circuit breaker
opens the circuit on the grid side. These time lags imply that
pressure is still building up, even though the rupture disk has
operated.
Further calculations made for the worst 236kA short-circuit
show that, with a 100mm, 4 inches, depressurization chamber,
pressure drops normally but by keeping the transformer
fed, the pressure quickly builds up again after approximately
54
Figure 17: Evacuated oil and gas volume versus Rupture Disk diameter
200ms because of the exhaust inefficiency.
Greater indepth research for every efficient depressurization
chamber diameters was therefore conducted to find the
minimum diameter to prevent such an occurrence from happening.
As a result, it was discovered that the minimum depressurization
chamber diameter had to be equal to or greater than
200mm, 8 inches, to avoid another pressure build-up after the
depressurization process, as shown Figure 18.
10. GENERATED GAS VOLUME
CALCULATION
Gas production has played a major role in the transformer
fire experienced by this western U.S. utility in one of their
underground hydro plants. The total plant, 3 units, was put out
of service for more than 4 months and the failed unit for more
than 10 months, because of the gases generated by oil during
the short circuit and resulting vessel explosion, but also by a
"fireball" phenomenon due to indoor transformer location.
When the walls inside a building are able to withstand an explosion,
the explosive gases do not have immediate oxygen to burn
instantaneously causing a "fireball" to travel, seeking oxygen.
In this case, the "fireball" moved inside the power plant gallery
to the 4 x 4 meter outer door, which was blown 60 meters away.
The company requested SERGI to calculate the gas produced
during the incident, in order to correctly size the TRANS-
FORMER PROTECTOR gas exhaust pipe that would be channeled
outside the powerhouse.
Figure 19 : Gas generation for the 35.4 kA fault
The MTH model and its associated software enable the
calculation of the amount of gas generated during arcing.
Continued on Page 56
Electricity Today