I recently experienced the failure 1

[tranny]

I recently experienced the failure of a 10 kVA constant voltage transformer utilized on the bypass side of a UPS that feeds a critical AC panel in a power plant. Here is what we experienced: At 11:00 AM someone smelled smoke from the SCADA cabinets. The doors were opened and a lot of smoke was coming from the power strips. (two power strips burned up, one tripped) Immediatly ran downstairs and found the 10 kVA constant voltage transformer (CVT) on fire. The circuit breakers were turned to the off position and the fire then subsided. Upon inspection of the CVT, it appeared that the inductor experienced a sigificant amount of heat from inside the core and coils, which then may have caused the insulation to fail. This may have then caused a voltage spike downstream causing a ground fault that wasn't cleared by the 175 A breaker.....for 10kVA at 120V, shouldn't this be a 90 A breaker anyway? (83.3 Amps) The fault current increased until stuff started catching on fire. I do not know if the system was in bypass when the failure happened, or if a failure at the power strip caused an overload on the inverter, which then caused it to switch to the bypass source. Can anyone provide some insite as to what may have happened. These constant voltage transformers utilize one or two inductor windings and capacitiors. Why are there two inductor windings, one bigger than the other?

 

[dydt]

hi..

please give us a clearer view how the CVT was introduced into the circuit..is it as follows?

utility - CVT - UPS (with static)

did it damage the UPS itself...can you transfer the inverter to static without problem..

CVT are pretty rough and don't go up in smoke easily...SOLA is one of the respected manufacturer of this....

..the CVT or constant voltage transformer also known as ferroresonant transformer...is transformer and by cleverly connecting the inductors and capacitors at resonant frequency (50 or 60 HZ) produces constant voltage..

dydt

 

[tranny]

There is one "essential panel" that is fed from two possible sources. Normally power comes from the inverter, through a static switch, to the panel. This provides clean ac power to the critical loads. If something happens, such as a fault, or overload, the static switch will sense this, and then instead of gating SCR's 1 & 2, it would shut these off and turn on 3 & 4. This will allow power to the essential panel from the Constant Voltage Transformer. The constant voltage transformer is always energized, but it's not under load unless the Static Switch goes to bypass.

The UPS did not fail. I do not know if the inverter failed because before I ever got on site, they called the inverter manufacturer and told them what happened. They stated that since the inverter was 13 yrs old, it should have maintenence. They came out and changed a bunch of control cards, and capacitors. They never tested it before replacing a bunch of parts. The CVT is a General Electric "Stabiltron", 10 kVA. Shouldn't this have a 90 Ampo breaker on the load side at 10 kVA and 120V?
Thank you for your help. I am new at this.

 

[dydt]

hi..

did any of the static switch SCR's shorted out?..1,2,3, and 4?

the capacitor replaced are they electrolytic, big diameter, approx. 3 inches? and how many did they replaced?

dydt

 

[tranny]

The static switch SCR's were fine. They replaced one capacitor in the inverter that they noticed was bulged. I think they are electolytic, big diameter, 3"

 

[tranny]

Here is a thought, see what you think.

When all this happened, the results were that two power strips were burnt, one tripped, and the Constant Voltage Transformer (inductor part) downstream was on fire.

Is it posible that the CVT was shorting internally, and that caused voltage spikes (V=L di/dt), and the the power strips were absorbing the voltage spikes until they had enough and started to get real hot?

And then it just so happened that the CVT primary breaker tripped in time so other equipmet didn't smoke. The secondary breaker wouldn't trip becaue it appears to be oversized at 175A instead of 90 A at 10kVA and 120V.

 

[dydt]

tranny..you ahve another posting about a UPS circuit..in this forum...are we dealing with same problem but different questions?

dydt

 

[dydt]

tranny..you have another posting about a UPS circuit..in this forum...are we dealing with same problem but different questions?

dydt

 

[tranny]

Yes, this is all part of the same problem..just different questions.

Does anyone know of any history regarding the General Electric "Stabiltron" transformer regarding possible failures of the inductor? Can you explain how the inductor is placed in teh circuit...I'm not quite understanding it according to the drawing. I wish I could attach the drawing.

 

[dydt]

hi..

as i mentioned on the other posting it seems that you used only one circuit of the CVT secondary, 120 volts, X1 and X2 instead of X1, X2 and X3 which is 240/120v center tapped.
the single 120VAC circuit was overloaded when you switched to bypass. I am sure that this is the first you used the CVT in that configuration via static switch. Yes, the essential load will run for a while while your winding starts to heat up. The secondary braker wont trip because your problem is not on the secondary circuit but on the CVT itself which is being supplied by your primary circuit breaker..

dydt

 

[jbartos]

Suggestion: The voltage waveforms could be analyzed for harmonic distortion and voltage harmonic content. The harmonics tend to heat inductors to a point of the thermal runaway.

 

[busbar]

Is it possible that the ferroresonant transformer’s secondary, rated “split” 120/240V 3-wire, is only intended for loading of ~42A X1-X2 and ~42A X2-X3; id est, 5kVA on each 120V-secondary winding?

Also, was there surge protection in the plugstrips intended for grounded circuits, and with the serving ferroresonant transformer secondary ungrounded—internal MOVs experienced overvoltage failure?

Third—compared to a similarly rated conventional drytype, this type of transformer generally has a comparatively high per-unit impedance. Relying on a molded-case breaker for secondary overcurrent protection may be effectively “cosmetic,” in that even an [NEC450] correctly sized secondary-side breaker may not protect the secondary from overload.

 

[nukeman]

It soounds like your CVT failed and caused the downstream faults. Check your grounding configuration of the downstream loads. We had a similar failure where a failure in the CVT caused a fire in a downstream UPS. The downstream fire was a result of a improper ground. As discussed in your other thred, grounding the output of the CVT is acceptable. To correct our design we ended up putting in fuses on the sencondary of the CVT, grounding the secondary of the CVT and correcting the ground problem at the UPS.

Our CVT had an internal lamination break down which allowed the core plates to break free and eat away at the winding insulation until a short occurred. Since this was on the secondary side of the CVT there was not enough current to trip the input breaker so the fault kept occurring until we had the two fires.

 

[tranny]

Busbar,
according to GE product information GEA-11062B, therecommended way to connect this thransformer for the nameplate kVA is to use x1 and x2. Using x2 and x3 will provide 1/2 nameplate, or 5 kVA. If it was hooked up this way, I can see there would be a problem. Also, the measured current (load) at the essential panel is only 35 amps.

 

[tranny]

jbartos,
I have suggested that a meter be connected to check the harmonic content. We will see how far the customer would like to go to investigate.

 

[tranny]

Could I get some recommendations as to how this system should be grounded. It appears that IEEE Std 142-1991 figure 10, configuration 2 is what we have, only its single phase instead of three phase.
The neutral that comes form the bypass transformer (CVT)is tied to the neutral at the UPS, which is grounded at the UPS.

Nukeman,
That is what I think happened as well. What was the ground problem you had at the UPS? How did you correct it?

 

[nukeman]

In our configuration the CVT was the normal power and the UPS output was the emergency power. The UPS and CVT were located in two separate rooms. If power to the CVT was lost the UPS would be cycled on. The original design had the ground at the UPS neutral and ungrounded at the CVT neutral. Poor design, normally you want to ground the source. We ended up removing the ground at the UPS and grounding the CVT neutral alnog with adding fuses to the CVT output.

 

[busbar]

tranny — regarding the Stabiltron transformer, are you stating that per the GE bulletin, X1-X2 has a full 10kVA rating, whereas X2-X3 has only 5kVA?

 

[tranny]

Busbar,
Yes that is exactly correct. X1-X2 has a full 100 kVA, and X2-X3 has 1/2, or 5 kVA

 

[busbar]

Interesting. The plain-vanilla symmetrical dual-secondary series/parallel arrangement is too ingrained in my habits, I guess.