Gassing in Furnace Transformer

[HoJoCo]

I have a customer with a 60 MVA National Industri unit that had circulating current problems that caused two of the top core bolts to be eroded away and break. All 8 core bolts were replaced and positively grounded. The unit was put into service in June-07 and is once again gassing at a high rate of rise (combustibles)
Hydrogen (H2):234
Methane (CH4):198
Ethane (C2H6):41
Ethylene (C2H4):334
Acetylene (C2H2):26
Carbon Monoxide (CO):33
Carbon Dioxide (CO2):332

I did an internal and have eliminated the core bolts/tap changer and secondary bus. Furans and metals are good!

Has anyone dealt with this type of problem as it relates to a furnace transformer? Any help/suggestions would be appreciated!

HoJoCo

 

[alehman]

Did you do an initial DGA? What were those numbers?

 

[HoJoCo]

We had just degassed the unit(8-25-07) the best we could: See below.
Samples listed left to right(9-13-07)(8-29-07)(8-25-07)
Hydrogen (H2): 234, 134, 0
Methane (CH4): 198, 110, 6
Ethane (C2H6): 41, 33, 7
Ethylene (C2H4): 334, 194, 46
Acetylene (C2H2): 26, 17, 4
Carbon Monoxide (CO):33, 17, 27
Carbon Dioxide (CO2):332, 213,163

 

[alehman]

Wow. Either there is an outside source of contamination or there is something seriously wrong with this transformer. These gases indicate serious overheating and likely arcing.

You really should be taking this transformer out of service before it takes itself out and you have a big problem and possibly people hurt. The transformer should be disassembled and thoroughly inspected to find the cause of this problem.

Also see

http://www.usbr.gov/power/data/fist/fist3_30/fist3_30.pdf

 

[unclebob]

I agree with alehman,

Acetylene is a gas that you don't wanna have.
It can only be produced by arcing.

Does the transformer has an on-load tap-changer?
Does the on-load tap-changer has a separate oil tank?
If not, you have a problem.

I guess the other gases are produced by the heat around the arcing area.

 

[HoJoCo]

The on load tap changers are in seperate tanks with a seperate oil system. We did an inspection of the tap changer and the main tank, finding no indication of arcing. I share the concern about Acetylene. I had hoped someone would have some experience in this type of furnace transformer (National Industri).

The unit is still gassing at a high rate!
Samples listed left to right(10-4-07)(9-13-07)(8-29-07)(8-25-07)
Hydrogen (H2): 497, 234, 134, 0
Methane (CH4): 416, 198, 110, 6
Ethane (C2H6): 82, 41, 33, 7
Ethylene (C2H4): 712, 334, 194, 46
Acetylene (C2H2): 65, 26, 17, 4
Carbon Monoxide (CO):52, 33, 17, 27
Carbon Dioxide (CO2):430, 332, 213,163

 

[DaveScott]

//All 8 core bolts were replaced and positively grounded//
Would that stop the circulating currents, or increase them? Were the materials correct?
The opposite direction would be, of course, to insulate them.

 

[HVDragon]

I made calculation of the DGA and the result is defect with high temperature 780 degrees celcius and discharges.
I didn't understand exactly what bolts were grounded. The bolts that press the core?

Something terrible is stalking between the high voltage potentials

 

[alehman]

I realize the economic implications of shutting this thing down are probably high, but so are those of not shutting it down and having a major failure (not to mention the hazard to personnel).

Then I suggest consulting the manufacturer.

 

[HoJoCo]

Sorry,
I was away for some time! The core bolts are insulated from the core as they pass through the clamping rings. THey are then intentially grounded at one end to cause them to be at the same potentail and eliminate circulating currents. This design was from the manufacturer.

 

[RRaghunath]

HoJoCo,

Could it be that the OLTC tank is leaking (or communicating) its oil into the main tank.

I had one situation, where the conservator was common for main tank and OLTC tank oils, then the separation between them was having weld problems.

 

[HVDragon]

My suspicions are confirmed. The core must be grounded only in one place to avoid the additional magnetic flux.

Something terrible is stalking between the high voltage potentials

 

[oldfieldguy]

On the "core ground" thing: A physical shock to a transformer, such as a heavy through fault or the mechanical motion of moving will sometimes create a second path to ground for the core. This second path will allow a complete circuit for circulating current and can result in some heating problems, including gassing.

Testing for this situation is sometimes nasty in that one must lift the manufacturer's core ground and megger the core against the case. There should be good insulation resistance with the core ground connection opened. If not, you have an inadvertent core grounding problem.

Occasionally the inadvertent ground can be 'burned' free by application of high current and low voltage to the core with the case as the return. If this doesn't work, though, you're looking at an untanking in most cases.

old field guy