Also ass-uming that fan cooling will make a difference. If it is rated 700 MVA with fan cooling then increasing the fan power will not make much difference. Once you get past a certain airspeed the cooling oil will not flow any faster.
Yes. Trnaformers may be loaded beyond the name plate rating either for short short time (ST) or long time (LT). However, beware that there is a risk of loss of life on the transformer and premature failure in and other components if this exceed the standard guidelines.
If the average ambient temperature is below 30 oC, the transformer can be safely loaded as per the ANSI Std C57 guidelines as shown in the curve below.
There are several loading practices in the utility environment. One of the this guide was presented in the study prepared by EPRI, IEEE & EEI based in statistics compiled In the pass 30 years showing not significant increase in transformer failure rates using the overloading criteria of 3% annual loss of life (ALOL) provided not exceeding the following constrains other component constrains.
mc5w... ass-suming you had read or seen the IEEE C57.91 standard, you would see that it has more to do with life expectancy changes for LT and ST overloading the transformer (and nothing to do with fan speed).
If your transformer was liquid filled and built to C57 standards, the standard relates to top oil(max 110C) and hottest spot (max 180C) conductor temps and provides a formula for computing loss of life expectancy for the transformer... the maximum loading referenced in the standard is actually 200% or 2 PU on liquid filled transformers with a standard 65 degree rise (but you get a loss of life for the transformer) and you have to be considerate of possible gassing within the transformer reducing the dielectric strength of the transformer oil.
By using the formulas you can compute the transformer life beyond or under the 20yr standard when you operate at different loads. I believe that CUKY2000s reference is for the avg 24hr loading which follows the C57.91 standard for no loss of life (and by the way was very helpful).
I was ass-uming that for this large of a transformer either the loading is 24/7 or that 200% overloading was not acceptable. The problem with 200% overloading is that the winding loss is 400% of 100% loading which for this power level would be terribly inefficient. Deliberate overloading is for the purpose of reducing no load loss by putting in a smaller transformer, not for the purpose of reducing capital cost. If peak demand lasts more than 3 or 4 hours because of air conditioning demand then you have a problem that the heat capacity of the oil and windings is not enough to help. Some of the "rules of thumb" for overloading of transformers were developed during the days when peak demand was due to early evening lighting use. If the transformer supplies both an air conditioning peak demand and an electric heat peak demand then deliberate overloading is even less acceptable. I have seen an electric heat neighborhood in central Pennsylvania where each pole mounted transformer serves 4 houses and is rated 167 KVA.
A formula is an equation that you use but do not understand.
Deliberate overloading of transformers really applies to the small transformers that service residential and commercial loads. In this case deliberate overloading is acceptable because a transformer failure only knocks out a relatively small amount of load. Smaller distribution transformer fires are also easier to fight. Also, smaller oil filled transformers are usually more robust relative to their size and the "full load" rating is really based on acceptable voltage drop when starting motors, not on what the transformer can really take.
This big of a transformer is either an important generator step up transformer or a transmission transformer that serves a major load. You can just as easily save on no load losses by shutting off the transformer during off peak periods and throwing load onto an adjacent transformer.
You are also not taking into account the extra reactive compensation needed to load this transformer at 200% of self cooled rating. The impedance of this size of transformer is around 10% at the self cooled rating to help control short circuits and to help paralled transformers carry equal load. 200% loading would require about 3 or maybe almost 3.5 times as much reactive compensation as 100% loading depending on how much no load magnetizing current that the transformer needs.
Pushing a 700 MVA transformer to 103% or 105% of maximum rating is "probably" acceptable, but you are increasing the chances of a major transformer fire. You would also need to pay more attention to keeping the radiator free of dirt coatings and debris. You would also need to have alarms hooked up to the overload relays for the fans so that you can fix a broken fans quickly - you would also need to connectorize the fans and otherwise make it easy to change out a fan.
You need to consider all of these other factors besides reduction of transformer life. If after computing the extra real power and reactive power to operate this transformer at a higher power level you might find that it is cheaper to add another transmission transformer and reduce the loading of the transformers. In this case, overloading of the transformer would only be needed if an adjacent transformer os out of service ( repairs, cleaning the radiator, yada yada )
leoliu,
While the preceding is all correct, I had a few experiences to add. Where I work we have transformers 30 years or more old. It has been found that the temperature rise on these in practice is much lower than specified for the insulation meaning that it is possible to get up to 20% extra out of them. The downside is that the transformer has to be dissasembled and every single bolt, conductor and lead checked to ensure that it is sized correctly and replaced if it is too small. Add some extra cooling and it seems to all go well. The main proviso is that you send the transformer to people who have the expertise in transformer design to accurately work our what is required, or with the courage to tell you that it can't be done.
There are also some oil manufacturer's who claim better cooling with their oil - another one to refer to the experts.
Are you measuring the oil temperature at the top of the transformer, the temperature of the insulation to oil junction, or the conductor to insulation junction? The conductor to insulation will be the hottest part of the insulation all other things being equal. The top oil temperature can be substatially cooler than the conductor to insulation junction.
You have to remember that there is a temperature difference across the insulation as well as a boundary layer between the insulation and the oil. There is also a temperature drop at the radiators that is essentially 3 layers consisting of the oil boundary layer, the radiator metal, and the boundary layer of air.
Once you get the air velocity and oil velocity past a certain point the boundary layers that impede heat conduction will not get any thinner.
Also, be VERY careful about doing a forced oil circulation refit. Moving oil ( or any other moving insulator ) can pick up static electricity. In other words, a forced oil design needs special shielding to discharge any static electricity bubbles that form because the oil is moving fast.
mc5w,
I agree with your thoughts and that is why I must refer back to those who have expertise in transformer design and manufacture to ensure that what is done is not cheating. In this case measured hot spot temperature during original testing was much lower than rated insulation temperature. This may have been due to substantial overspecification when originally built. Some of the increase may also be due to improved design methods.
You could also have an oil viscosity difference that slightly improves cooling when the weather is warm.
Also, the heat transfer characteristic of a solid to fluid boundary layer can be quite difficult to estimate or measure correctly, so I can agree that transformers are slightly overdesigned. Maintenance procedures such as cleaning radiators, checking the fans for correct operation, and so forth can make a big difference.
I am also suspicious that different transformer fluids have different cooling performance and transformer manufacturers design transformers for the least performing fluid that could be in the transformer.
