Transformer gas analysis 1

[eeprom]

I have some transformer gas results which show large increases in H2 (>300%), CO(15%), CH4(200%), C2H6(120%), C2H4(30%). Tests were taken 3 months apart. There is not enough motor load on this transformer to overload the transformer. The transformer is 5MVA. The largest single load is a fan of about 1.25 MVA. Between take of oil test 1 and oil test 2, the large fan was balanced. And in doing so they had some problems, so the fan was allowed to start many times in the course of an afternoon. I don't know how many times, but I'm sure they didn't stick with the 2 starts per hour rule. The fan takes close to 40 seconds to start. Of course the starting current is some 600% of the normal current. I don't know if any other loads were active when the fan was being balanced.

Does anyone know if multiple starts of this large fan would cause enough internal heating to result in such a large change in gases?

thanks
EE

 

[prc]

What is the absolute values of gases in ppm before and after the incident?

 

[electricpete]

Does anyone know if multiple starts of this large fan would cause enough internal heating to result in such a large change in gases?

1 – I agree with prc. Tell us ppm.
2 - fwiw, assuming you are seeing a significant increase in gases such as H2, CH4, C2H6, C2H4 different than historical, my guess would be no, these are not related. The “two starts per hour” rule is undoubtedly to protect the motor. 40 seconds is a long start for the motor, but is just a short load blip for the transformer. (I don’t think the transformer cares that the load turns much more inductive during this period... open to comments). If anything you might see tiny blip in CO and CO2 from the increase in temperature, but not those other gases.


=====================================
(2B)+(2B)' ?

 

[eeprom]

I agree that the excessive motor starting is not likely to cause the problem. I think that the transformer could handle more than that before overheating. But the motor starts might aggravate a problem.

I incorrectly stated the results, so the numbers I mentioned in the first thread are not accurate.

The tests were 4.5 months apart.
Gas ppm1 ppm2 2 yr trend IEEE state
--------------------------------------------------
H2 329 --> 1107 Increasing Condition 2
CH4 35 --> 69 Increasing
C2H6 11 --> 19 Increasing
C2H4 12 --> 12 Increasing
CO 500 --> 451 Increasing/flat Condition 1
CO2 7049 --> 8914 Increasing Condition 2
TDCG 886 --> 1658 Increasing Condition 1

TDCG/day = 3.5 ppm/day

ppm1 are results from 4.5 months ago
ppm2 are current results
2 yr trend is generalized trend since 2009
IEEE state comes from IEEE STD C57.104-1991

I think that the presence of the high energy gases are occurring due to a localized heating, most likely at the load tap changer. I think that there is a high resistance spot on the tap changer and the frequent motor starts provide a short term heating of that spot. This is where methane, ethane, and ethylene are coming from.

 

[electricpete]

likely at the load tap changer.

Is the load tap changer in the main tank?

Our LTC is in a separate compartment (not familiar with LTC in main tank). We have recently seen an increase in LTC compartment gases associated with change in load. The change in load caused the LTC to switch to an infrequently used contact and gasing only occurred when on that contact.

=====================================
(2B)+(2B)' ?

 

[electricpete]

This thread is where we saw gassing only in one position of LTC:
thread238-304522


=====================================
(2B)+(2B)' ?

 

[electricpete]

So, just to finish the thought. It may be that motor starting causes your LTC to switch to a position where gasing occurs (assuming these results are from the same compartment as the LTC).

That's not too much different than what you were saying.

=====================================
(2B)+(2B)' ?

 

[DTR2011]

I would guess that for a 5 MVA unit, the OP is referring to a ULTC.

A winding resistance test on all tap positions may help identify the problem (what phase, likely the connected tap). A TTR may be of some assistance. This could also be a loose internal connection.

Nothing beats an internal inspection. Depending on where the No Load (ULTC) tap changer is located, the problem may be easy to identify.

In larger units (from Europe), I have seen internal LTCs, quite common in fact.

 

[electricpete]

In my defense, I read LTC and assumed it meant LTC.
But you’re right, that’s not likely an accessory for that small a transformer.

=====================================
(2B)+(2B)' ?

 

[eeprom]

Excuse my vocabulary. It is a de-energized tap changer, used for moving voltage up/down in 2.5% increments. It is on the main tank. As far as I know it has never been changed in 20 years. An internal inspection would be ideal, but it would also be very expensive.

 

[stevenal]

What's the C2H2 doing?

 

[stevenal]

The most recent C57.104 is 2008. Using this standard puts your transformer in condition 3 based on the H2 and CO2 levels.

 

[prc]

Please check the trend of oxygen.If it is coming down, coking at copper contacts is to be suspected. Since there is no C2H2(acetylene)I am not suspecting such a case. Overall I believe,there is some PD,no heating anywhere. Watch for any increase in gases by monitoring gases everymonth, till gases stabilises.Also the H2 generation can be stray gassing. IEEE Std was revised after 1991.Refer to new standard.

 

[stevenal]

prc,

Please define "stray gassing."

 

[eeprom]

PRC,
Very interesting insight. The O2 has gone down quite a bit, from 2953 to 1876. The CO2 has gone up from 7049 to 8914. The H2O has gone up from 57 to 68. I would assume that there is some combustion happening in the oil, which implies localized heating.

By the way...what do you think IEEE means by a partial discharge? How can a discharge be partial?

 

[dpc]

Partial discharge is localized breakdown of the dielectric. It doesn't fully bridge between the two conductors. Similar to corona in air.

 

[eeprom]

So then where does the discharge go? All current flow needs a circuit.

 

[ScottyUK]

eeprom,

PD is a problem in solid dielectrics. It doesn't affect liquids as these are self-healing and voids don't occur. Have a look at the wikipedia entry for an intro. It's a complex subject, and people spend their careers studying it.


Off thread a little: tapchanger terminology:

- OLTC (on-load tapchanger): designed to change taps while carrying rated current and live at rated voltage.
- OCTS (off-circuit tap selector): designed to change taps with the transformer windings made dead and isolated from all sources.

I have never seen an off-load tapchanger, a design which presumably allows the tap to be changed with no current flowing but live at rated voltage. Can anyone point to an example?

Pedantic? Perhaps. Three colleagues - friends - died because they operated an OCTS while the transformer was energised. They believed, incorrectly, that it was an off-load tapchanger. The resulting fire and explosion totally destroyed the transformer. The lads were caught up in the fireball.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[rterickson]

The term over here used to be NLTC - No Load Tap Changer, but now we use DETC to emphasize De-Energized.

 

[dpc]

Scotty,

The term "no-load tap changer" is extremely common in the US and has been for entirety of my(nearly) 40 year career. It means a tap changer intended for de-energized operation, but I agree that it not strictly what the words seem to mean. It's probably time to retire the term. I'd never even heard the term "off-circuit tap changer" until a few years ago.

In almost every instance I can recall, the tap changer operating handle is well-marked with a sign indicating that it should not be operated while the unit is energized.

This is not meant in any way to minimize the tragedy you described, but just to say that this terminology is in wide use, at least in the US. In the past few years, the term "De-energized tap changer" or DETC has become more common. Probably due to the ABB DETC mechanism being used in most US transformers these days. I agree that this terminology is much better than "no-load tap changer".