Problem with transformers on cogeneration plant 1

[PMACP]

Hi,

I need help to understand a problem I have to solve in a cogeneration plant. The power transformer that do the conncetion between the generator and the network has problems which I can't understand. That transformer fails because it heat excessively.

Some help?

Thanks
Paulo

 

[PMACP]

alehman:

Now, the photos. Sorry.

Thanks

 

[alehman]

Have you had these transformers inspected by the manufacturer? Has any testing been done? Is there a permanent short? if so, where?

I don't think flashover at that location would have been the result of overheating. The most likely cause would seem like a voltage transient from switching operation. Based on the burn marks along the vertical bars, it probably started near the 90 degree bend just above the coils. The space between phase and neutral looks rather small. In the U.S. 1 inch (25 mm) would be required.

 

[PMACP]

alehman:

Hi. The manufacturer inspect the transformers, and point out some possible causes, namely: harmonics; transformers working in a permanent overvoltage, possible saturation of the nucleos, and the temperature of the room.

In my measures I don't see any problems of harmonics (THD-I < 2% e THD-U < 2%). The voltages in the primary are in the range of 403 V to 435 V for a rated voltage (on primary of the transformer) of 400 V. I think that overvoltages are not sufficient to justify saturation. What's your opinion?

The room temperature, at the time the transformers were destroyed were 40-50ºC, wich is significant. Is that possible that the room temperature has weakened the insulation? Moreover, the overvoltages may also contribute for that.

The defect you refer could ignite the transformer? As far as I know, one of the transformers was in fire and the other only was cooked.

The voltage transientes are a good possibility, but i need more details to understand that. One of the transformers had the problem two hours after the generator was switched on (when the transient on voltage could have occurred). It is possible that the transient has had this behavior?

Thanks

 

[edison123]

The transformers see voltage transients/surges all the time and are designed to withstand them. That's why they are routine tested for twice the rated voltage and type tested at many times the rated voltage during an impulse voltage withstand test. 50 deg C ambient is not such a big factor here especially when the transformer was being loaded only to 80%. The manufacturer doesn't seem to have the guts or the honesty to admit that his transformers are poorly designed and made and is trying to shift the blame to the operation side.

 

[PMACP]

edison123:

Thanks for your opinion. Please, can you give me some more details about your tip? Note that the transformer datasheet refers a maximum ambient temperature of 40ºC, imposing also a daily average temperature of 30ºC and 20ºC for annual average temperature. In fact, the transformers were operated above those values (in fact, above the maximum permitted). Also, in the datasheet we may found that for an ambient temperature of 30ºC (anual average), the transformer should be derated in 10%. If we assume a linear derating process, this means that we have a 15-20% derating factor for the present situation. You think that this kind of issues were not sufficient to damage the transformer?

Thanks
PMAC

 

[edison123]

Answer to your last question - No, I don't think "the issues" that you talked about here, at length if I may say so, are not, repeat not, sufficient enough to destroy a trafo.

Please read the transformer standards about the temperature limits and what they mean.

 

[alehman]

There have been an increasing number of flash-over failures in dry-type transformers caused by switching transients. IEEE has formed a working group (WG C57.142) to investigate this problem and make recommendations. So far, no standard has been published. These cases mostly involve vacuum or SF6 interrupters and short cable runs.

I agree that 50C ambient with 80% load should not cause immediate failure.

 

[WhiteyWhitey]

PMACP

Has there been any estimate of the magnitude (Size) of the switching transients?

Is there any chance of a transient reflection at the insterface points of the cable and could this have contributed to a higher transient than the transformers surge rating.

Are there any surge protectors installed on the Transformer or Sub?

Cheers,
Drew

 

[PMACP]

Hi,

edison123: I read the standards about the temperature limits, but its meaning is not clear. They set 3 limits: anual average ambiente temperature (20ºC); daily average temperature (30ºC) and maximum temperature (40ºC). All that values refer to the cooling fluid, that in this case is the air. There's no information about the meaning of those values. Obviously, I think the meaning and the second temperatures are clear, but not the last one. What means the maximum temperature? How long a transformer may work at 40ºC ambient temperature? That maximum must not be violated? or if it is violated only a reduction on life expectacy od the transformer occurs?

Is there someone that has the standard IEC 76-1 to IEC 76-5?

Thanks
PMAC

 

[edison123]

Hi PMAC

Since the ambient temperature variations are not within our control, the standard specifies multiple limits from a maximum of 50 deg C to maximum daily average of 40 deg C to maximum yearly average of 30 deg C to even a minimum of -5 deg C.

Per standard, the transformer is supposed work continuously at rated load with a normal design life expectancy (which is around 20-25 years)

In your case, 50 deg C ambient should not destroy the windings prematurely especially it is only loaded up to 80%. The temperature is the major killer of any winding (motors, generators and trafos) and hence it is imperative you have a working temperature sensor/protection (even a redundant one) to monitor and save the trafo.

Assuming yours is a class F, dry-type trafo, the maximum winding temperature allowed is 130 deg C with a normal design life. Have your trafos ever recorded temperatures beyond 130 deg C ? Did you witness the factory acceptance test of these transformers where they are supposed to do temperature rise test ? If yes, what was the maximum temperature in that test ?

 

[PMACP]

Hi edison123

Thank you again for your help. First of all, I would say that the transformers had a Z protection against excessive temperature, but them didn' work.

"...the standard specifies multiple limits from a maximum of 50 deg C to maximum daily average of 40 deg C to maximum yearly average of 30 deg C to even a minimum of -5 deg C."

OK. I don't have the standard (IEC 76), so, I don't know how should I derate the transformer in order to avoid excessive heating. Can you help me on that? The maximum value os maximum temperature defined on the standard is 50ºC for a average daily temperature of 40ºC and average annual temperature of 30ºC?

"Per standard, the transformer is supposed work continuously at rated load with a normal design life expectancy (which is around 20-25 years). In your case, 50 deg C ambient should not destroy the windings prematurely especially it is only loaded up to 80%."

Well, if the transformer works some months with maximum temperature of 50ºC, the life expectacy is reduced, right? How can I quantify that?

"The temperature is the major killer of any winding (motors, generators and trafos) and hence it is imperative you have a working temperature sensor/protection (even a redundant one) to monitor and save the trafo."

That's an interesting issue. The Trafo was protected using a Z protection with 6 PTC, one for winding. Those protection, as long I know, doesn't work...

Moreover, there are the arcing effect that I presented before (you can see the photos on a preview post in a pdf document - 14 Feb 09 4:21 ). On yours opinion, what may be the reason to that arcing effect?


"Assuming yours is a class F, dry-type trafo, the maximum winding temperature allowed is 130 deg C with a normal design life. Have your trafos ever recorded temperatures beyond 130 deg C ?"

In fact it is a class F transformer, but I don't have any information about those temperatures.

"Did you witness the factory acceptance test of these transformers where they are supposed to do temperature rise test ? If yes, what was the maximum temperature in that test ?"

I don't have that information. I have information for another transformer of the same manufacturer, but this kind of information doesn't appear. However, I think it should be in accordande with the standard, right?

Thanks

 

[edison123]

You should not be operating those trafos without thermal protection. Fix them first pronto.

You have already derated the transformer (either by design or by accident) to take care of 50 deg C ambient. So again, this is not a cause of your trafo burn-out.

Arcing is probably due to insufficient spacing, which leads us back to poor design/manufacturing.

As I said previously, you need to post some hard numbers here to go any further. Otherwise, both of us are just wasting time.

 

[waross]

Is it possible that the burn-out produced smoke, soot or flames that may have ionized and started the arc?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[PMACP]

Hi,

I'm finishing the collection of data on two transformers. Today I have some data that may help understanding what could happened. However, I have some questions for which your help is welcome:

a) When a transformer saturates, there are harmonics in the currents of magnetization. The fifth harmonic in particular, is often used to detect situations of saturation. What amplitude of those harmonics should indicate saturation?

b) What level of overvoltage could lead to saturation of the core? 5%? 10%? 20%? ... For instance, a transformer with a secondary voltage adjusted to 15400 V could be inm saturation when it is subjected to a voltage of 16500 - 16750 V?

Thanks

PS: I'll put some data here en the next days.

PMAC