Brief removal of cooling on generator stepup transformer ok? 6

[electricpete]

Generator stepup transformer 25kv/345kv Shell Form 650MVA FOA rating. No FOA or FA rating.

We are studying POSSIBLE options to address identified thermal hotspots in the power supply to the fans/pumps. Someone had proposed removal of all cooling power for a brief interval to allow repairs. During this period the transformer would be fully loaded and have no pumps or fans operating.

We are trying to determine what duration if any would it be safe to operate like this.

My concern is that with oil flow greatly reduced (only small amount of natural circulation), we have no ability to determine actual winding temperature. (delta between oil temperature and winding hot-spot will be much higher than it would be with flow).

What do you think? Has anyone ever done it?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Additional fans is easy.

There is no practical way to rig up additional pumps while the transformer is running since that is intrusive on the oil system. But perhaps as someone else suggested wire alternate power to some of existing pumps.

At this point as I said before I am most interested in just understanding the answer to the question - how long do we have before damage occurs IF we encounter a situation where power is inadvertantly lost.

GamezBeCJ - what type of transformer was that? Do you know roughly the MVA rating and voltage? Shell or core form?



=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[ruggedscot]

How long have you got - Id say looking at it and the design of the transformer you would not have long - if the oil flow stops you get very quick localised heating - same as if you have a blockage, the area would develop a hot spot leading to oil gassing and breakdown - probably take out the protection and drop the load. best case. worst case would be catastrophic failure.

http://www.powerlineman.com/downloads/download.html

go have a look at the video file called

Transformer Failure.mpeg

Now that is something that you really really dont want to provoke - as I said power xformers need to keep what they were designed to operate with working, if not then you fall down very very rapidly.

Rugged

 

[prc]

From the discussions ,it is a large transformer with directed cooling capacity only.It means oil is directed in to winding and guided.It is better you consult the manufacturer (as he knows all the weak spots in it )and he will be able to give a number to you.Normally this is checked like this .Take the full load winding hot spot temperature.Then assume that the entire winding copper loss (heat )is absorbed by the copper .Then find the time for copper temp to reach 140 C from the starting temperature ie full load condition.(usually less than 98 C )
Usually this figure is 10- 20 minutes.Be cautious ,as when winding temp goes over 98 C ageing rate goes up and chances of gas bubble formation and consequent failure is more esp if there is lot of moisture in paper .(You can ascertain this by checking water content in oil when trf is under max hot condition ) If the trf is old and water content is pretty high ,limit hot spot temp to 120-125C .You can also save the situation by undertaking the operation when ambient temp is as low as possible and load is as low as possible.

 

[electricpete]

Thx PRC. That sounds like a parallel analysis to what I suggested above, except with a more conservative assumption that heat stays in the copper, rather than heating up the entire transformer mass uniformly. IF we could neglect heat transferred to the environment, then the actual response probably lies between the two.

My transformer outline drawing indicates that the weight of the “core and coils” is 752,000 pounds. What fraction would be copper?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Now that I think about it, a better way to say it (rather than saying the actual response lies between the two IF we ignore heat transfer) is that your approach is conservative. I like that approach. Still need to know what fraction core and coils is copper.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[prc]

electricpete,sorry ,I will not be able to guide you to correct number for copper as I have no experience with shell type transformers.But your assumption of heat distributing to entire inside mass inside transformer is not correct as the time constant for such heating is more than 2-3 hours while we are considering a time frame of 10-15 minutes .So the conservative calculation will be safe.Of course a finer calculation can be seen in the IEEE/ANSI loading guide on transformers.

 

[GamezBeCJ]

Eletripete:

Sorry it took so long. Usually in America you will se transformers with that kind of cooling (OFAF or ODAF) starting from 100 MVA up to 500 MVA. The half hour I mentioned is a little bit conservative, since this is considering that the top oil rise is at is rated value (say 65 celsius). If the unit would loose cooling while partially loaded, you would have some more time before you start causing excesive aging to the insulation system. I would say that at the most you would have about 3 hours.
Regards.

 

[GamezBeCJ]

One more thing. As it has been sugested before. Contacting the manufacturer is a very good idea. He has all the calculations and models needed to predict the cooling capacity of the tank alone, and would give you a much more accurate estimate.

 

[Walden]

The Russian power station engineers removed the cooling for a short time at Chernobyl. Look what happened to them!

 

[itsmoked]

Can't ask them can we...

 

[electricpete]

for the record I am only interested in responses that have some technical basis related to transformers

opinions and analogies are worthless to me


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[abeltio]

there are on-line devices that can check for degradation of insulation (release of aromatics, gases, CO2, etc)
these devices (there are quite a few in the market, i would refrain from mentioning any brands) can be installed while the unit is in operation.

after obtaining a baseline... the instrumentation can be used to monitor the oil for any symptom of decomposition of the oil and/or the insulation.

in any case... removing the pumps/fans from service with the unit in operation would require by-passing some protections and trips... in Europe this is against the law, and voids any insurance, warranty and creates criminal liabilities...

in the us. i do not know... but at least i would check the insurance policy... if there is a conscious decission to bypass protections... it may also amount to criminal liabilities... even if nothing happens, the plant logbook will have to log the by-passing of the protections... so it will be subject to scrutiny by the cognizant authority and create a liability... something like "reckless driving".

if the risk of not doing the repairs on-line is high... the risk of doing the repairs without disclosing the situation to all parties is even greater in $$$
if push comes to shove... i would either schedule an outage or... bring all parties and explain the situation and justify the technical case for by-passing protections... this may include:
cognizant authority
neighbours representative
labor unions
customers
owners
fire marshall
OSHA
etc.

HTH

saludos.
a.

 

[electricpete]

We have a Hydran monitor installed on this unit. I don't have much confidence that the response from that unit in the absence of oil flow would be very fast. Also if hydrogen is not generated, the sensitivity of that device is fairly low (lower response to CO and CO2 than to H2). Surely if we got an increase in Hydran ppm indication that would be incentive to do something immediately but I wouldn't count on the device as a means to guarantee we're operating within a safe region any more than I would count on the temperature indicators (as discussed above delta between winding hotspot and top oil would be much higher during low oil flow).

I have never seen a generator stepup transformer with a trip upon loss of cooling. Have you abeltio?

Although not conclusive, that fact would seem to argue that the protective relay folks think the transformer can withstand momentary loss of cooling.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[abeltio]

what i have seen is that the cooling is controlled by temperature switches, which in turn send signals to the dcs...
there are two levels alarm, and normal shutdown... cannot recall the settings.
and of course the buchholz relay.

in one sad incident... the buchholz actuated, the operators thought it was a relay glitch... bypassed the protection and had a very hot copper soup in their hands in less than 10 min
they are still waiting for a new transformer.

with this, i mean that if there is a fault... it escalates to catastrophic failure very quickly.

hth


saludos.
a.

 

[GamezBeCJ]

Electripete:

Do you have to service all pumps and fans at once?, can you do one by one?.
You could de-energize one device at a time, try to reduce the load on the transformer to the minimum possible and keep an eye on winding temperature indicator.
If you do this during winter time, it is even better.

 

[prc]

electricpete,none of the hydran will be of any help to you in your plan of action.The failure mode ,if it happens will be sudden ,probably you will get no time .However the only item that will give some indication will be gas operated relay .(Do you have one ?)This will give an alarm or better watch for gas during the period of operation ,if you have brought the gas sampling device at ground level.

You need not by pass any protection device.Winding temperature indication will not go up as some of us suspect.It is a slave operated by winding current.Winding current is not increasing during cooler switch of mode .Oil temp will not go up as time of cooler shut down is much less than oil heating time constant.

If you can give me moisture content in oil with trf oil temp (at the time of sampling ) -take a sample when oil temp is max -I can help you in finding the water content in paper .If it is less than 2% ,we can safely reach a hot spot of 130-140 C.From test report pl find out the hot spot temp rise reached .Then you can calculate the time by assuming a copper wt.It will not be far out compared to a core type trf.Pl find the copper losss at fuul load from test report
Pete,I did not understand your problem fully.What is meant by "hot spots in cooler power supply "

 

[sparky1976]

Pete,
//Additional fans is easy.

There is no practical way to rig up additional pumps while the transformer is running since that is intrusive on the oil system. But perhaps as someone else suggested wire alternate power to some of existing pumps\\.

Eventhough you not interested,
the pumps is to circulate oil, so if you have mobile oil purification plant that can do online, I think you can do oil circulation with purification plant with the same flow/gpm specs. you can do this with spare pumps for your xfmr from your ware house. just be sure you remove the air from temporary hose or pipes.
I donot know if your xfmr have valves connection for online purification.
This if you decide do not want to loose the pump duirng transistion of supply to temporary supply.