Power Transformer Emergency Overload Condition: Keeping Temp Low 1

[111R]

If there's a situation where a transformer must be loaded 125% or more above the top ONAF rating for a short period of time, what are some options for preventing excessive temperature rise?

Aside from additional fans, has there been any experimentation with other ways to provide additional cooling that does not require taking the transformer out of service? For example, would multiple large industrial air conditioner units blowing directly on the radiators have any effect on an outdoor transformer? Or, some sort of water mist cooling unit?

I realize these are a bit off-the-wall, but I was wondering what options are available if you're ever in a bind.

Thanks

 

[mgtrp]

We used to run a weeping hose through the radiators to keep a troubled transformer down to a more acceptable temperature on hot days. It worked quite well.

 

[oldfieldguy]

Assuming that you keep water off the bushings and terminations, a flow or mist of water will carry a lot more heat away than any reasonable airflow.



old field guy

 

[davidbeach]

Why worry about the bushings and terminations? They're exposed to the rain on a regular basis.

 

[racookpe1978]

Use clean water or tap water. Not river water or cooling water contaminated by particles and other residues/animals/barnacles.

 

[FPelec]

According to IEC transformers loading guide (IEC 60076-7 ), 125% should be a feasible cyclic overload, even for large power transformers.
However, if the transformer is a large and new one, it could be a good idea to foresee an integrated real-time temperature / dissolved gas monitoring system.


Si duri puer ingeni videtur,
preconem facias vel architectum.

 

[oldfieldguy]

davidbeach-

That's fine for outdoor-style bushings and a droplet or mist application. I've been told by fire-fighting 'professionals' that the stream from a firehose is safe to 69 kV. However, none of them accepted my offer to view a demonstration.

I remember, however, a client's transformer that suffered a poor foundation, allowing it to tip to the point that the upper tubes of the radiators were no longer in oil. This killed the natural thermosiphon effect through the radiators. The transformer heated up rapidly.

It was the sole power source for critical environmental equipment and could not be removed from service, so we cooled it by flowing water over the case. The high voltage (13.8 kV) bushings were inside a termination cabinet, a fortunate thing because the long-term flow of water over the case resulted in long beards of algae growing from the transformer. Clearly, as a long term solution, this had some issues. That transformer was still in service with its long green beard when I moved to another service area.

old field guy

 

[davidbeach]

Ours spend months every year being drizzled on (one of those places where "it rains all the time") and I'm not aware of algae ever being a problem so that never crossed my mind. We have a few where we'll set sprinklers on the hottest days of the year, but in a bad year they might run 96 hours.

 

[krisys]

Please note that the additional heat will be generated from the transformer winding for any load over its 100% rating. The additional heat flow from one point to the other in a given medium is possible, only when the temperature difference between these points increase. (It is analogous to the Voltage – current relationship when current is flowing through a resistance). Thus in order to push the additional heat from the winding, the winding temperature has to be higher. This would give rise to an accelerated deterioration of paper insulation which is directly wrapped around the conductor.
Every time when the insulation temperature exceeds its rated value, its life reduces and the reduction in the insulation life keeps on accumulating. Hence in my view cooling the transformer as proposed by “111R” shall not be advised by an Electrical Engineer.

I have seen the mechanical guys in a process plant cooling the heat exchanger plates by continuously spraying water.

 

[prc]

Transformers are designed for overloading and the general rules are given in following over loading guides-IEC 60076-7:2005 or IEEE C57.91-2011. A study of these will clear the query.

The rating of transformer, age of it and period of overloading in hours and ambient at the time of overload will make the estimation easier. As per IEC Power Transformers can be overloaded (normal cyclical loading) up to 130 % provided,absolute temperatures of oil/winding and winding hot spot do not exceed 105/120 &140 C. The understanding is that loss of life at the time of overloading is compensated when the unit works under load or below normal average ambient temperatures.

For every 6C rise above the nominal winding hot spot temperature of 98C ( based on weighted annual ambient of 20C and max ambient of 40C as per IEC 60076-2) the paper life halves.

External additional cooling by way of extra fans or showering with water are all good to reduce the oil temperature rise, but still the winding rise will be more than normal and to that extent additional ageing must be expected.

Another point is that if the transformer is old, it is dangerous to overload the unit. Paper insulation in old transformers will contain 2-5 % of water in it which will be vaporized at temperatures above 100 resulting in bubbles in electric field which is a sure way for dielectric failure. Anyway a complete insulation dry out before the overloading cycle is recommended. When moisture exceeds 1 %level, insulation ageing goes up multifold for the same insulation temperature.

 

[jkristinn]

prc says:

For every 6C rise above the nominal winding hot spot temperature of 98C ( based on weighted annual ambient of 20C and max ambient of 40C as per IEC 60076-2) the paper life halves.

The underlying physics are based on Arrenhius's equation (which describes the relation of a rate of a chemical reaction to absolute temperature, the chemical reaction being the degrading of the paper in our case). As such, it is the absolute temperature of the hot-spot that is important for the insulation aging, nominal temperatures, winding gradients and temperature rises are only important in relation to the absolute hot-spot temperature.
Therefore anything that will lower the absolute hot-spot temperature will have positive effect on the degrading of the paper. Including lowering the ambient and adding thermal convection from the radiators (which will result in lower bottom oil temp and thus lower hot-spot although the temp rise stays unchanged).

 

[waross]

External cooling may keep the hot spot temperature down. That is why there is a difference between ONAN ratings an ONAF ratings. The added cooling reduces the oil temperature at a given loading. This cooler oil in turn holds the hot spot temperature down.

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

 

[prc]

The 98 C temperature mentioned by me is the absolute temperature and I never said anywhere that it is the rise.The standard (Both IEC &IEEE) take a nominal hot spot temperature for 'normal' life. It was proved by experiments that in case we keep paper temperature at 98C continuously, the life will be around 18.5 years. But we know that in reality, during service, some times temperature will shoot out above 98C due to high ambient temp or overloading) and many other occasions it will be less than 98C. So what ever loss at above 98 C will be compensated at when it is less than at 98C

 

[Compositepro]

I would like to point out that the fact that the normal boiling point of liquid water is 100C, is often a red herring in many technical analyses. The pressure in a transformer, under oil, is greater than normal. Also, there is moisture in paper because there is an affinity between the molecules, which lowers the vapor pressure. Nucleating and growing a bubble can require significant pressure above the vapor pressure. This is just a reminder that under these condition there is no longer any real significance to 100C, which is only relevant to liquid water at one bar absolute.

 

[cranky108]

Compositepro, you are assuming sea level. At higher elevation the BP of water is lower. But you are correct that PNB is a more likely boiling, and is random.

Depending on the transformer, it maybe at outside conditions, or in maybe under a N2 blanket. In either case if it is of signifent size it would have a chenical dryer attached.

The small transformers may not have much at all, and may be consitered expendable.

It is possible to increase the number of fans, cooling fins, add water cooling, but that dosen't change what is happening at the core. The oil can only draw heat away so fast.
What I saw once was to leave the cooling fans on all the time to allow the transformer to cool to an extra amount at night, so it stayed cooler during the day.

Really, water cooling? How many substations have water plumed to them?

 

[rterickson]

Pretty common for hydroelectric GSU's to be water cooled.