Transformer loading

Status
Not open for further replies.
B

Bjorn Larsen

Member
Location
Seattle Area
I have a 300Kva Aluminum wound transformer that has a continuous load of
259Kva (86%). Does anyone know what the "MTBF" (Mean time before failure) is for this situation?
 
Depends on ambient temperature, ambient conditions and temperature rating of transformer. If it is in a clean 35C enviroment with no vibration, no moisture, no excessive voltage spikes or inrush currents, MTBF is proably 25-50 years.

Theoretically, it could run at 300 KVA for 30 years.
 
Transformer is located in a UPS/electrical room, no dust, moisture or extreme temps. But isn't it good design practice to keep the load under 80% of the transformer rating.
 
Transformer aging is directly proportional to heat, so the lower the temperature, the longer the life of the unit. So, any load lowers the life of the unit.
However, 86% loading is within the nampelate rating, and with it being located in a date center within a controlled environmment, I wouldn't worry about the unit.
Critical Power engineers tend to use the 80-90% level as a redline value for the electrical equipment, but the transformer is designed to run at 300 kVA continuously - so I wouldn't be worried about it running at 86%.
One strike against it is that its aluminum.
I suggest perhaps an infra-red scan of the windings and terminations to give you some degree of assurance.
JM
 
I did suggest to IR scan the transformer, but I need to de-energize it first, due the interlocking cover, GE transformer. The owner will not allow this to happen, since we then woud be dropping the data center.
 
I was going to ask i you wanted that MTBF is years or decades.:smile:

My experience with transformers if they are kin to cock-roaches in terms of life. I've worked around many for years and years now that are hot enough to fry egges 24/7 and they keep on tickin'
 
You failed to advise what the temperature rise and what the insulation class was. Another assumption is that is is a general purpose ventilated transformer (xfmr) installed with at least 6" of ventilation space around it.
It is most difficult to state an opinion with out knowing all of the parameters.

Today?s 220deg insulation is far superior to those in the past for insulation life. A transformer manufactured with today?s material should have a 25-40 year life if the insulation class of 220deg C is not exceeded
It starts with the insulation class, i.e. 220degC. Then xfmrs are designed to operate at 30degC ambient, 40degC max. The maximum temperature allowed within the winding, commonly called hot spot temperature, is 30degC.
This is how it works:
40 oC maximum ambient
+150 oC maximum average winding rise
+30 oC maximum hot spot in winding
=220 oC ultimate temp at hot spot which is the insulation class

Overload a 220degC insulating class xfmr will have an affect on xfmr life by increasing the temp rise. Lowering the ambient temp can counteract this. As you can see you're dealing with amb. temp which could very and xfmr loading which can also very. On larger xfmr s where fans can be added (FA) the KVA can be increase 33% if the 40degC ambient is not exceeded.

Now, consider xfmr with 220degC class insulation but with a 115degC rise. 40degC+115degC+30degC=185degC that's 35degC less than 220degC. Xfmr s designed with a 115degC would certainly last longer because at full load they would run cooler.

Specifying 220 oC insulation with 80 oC or 115 oC rise will provide overload capability, better efficiency, as well as longer life. These transformers are designed to operate with a lower rise. As and example: 40 oC maximum ambient
40 oC maximum ambient
+80 oC maximum average winding rise
+30 oC maximum hot spot in winding
=150degC
220 oC ultimate temperature at hot spot - 150degC = 70degC that you have left to place with, i.e. overload capacity without exceeding the insulation class of the xfmr.
In other words a xfmr with 220degC insulation with an 80degC rise will allow for a 30% overload, a xfmr with a 115degC rise allows for a 15% overload keeping in mind that the 40degC ambient can not be exceeded.
Now, how do you predict voltage surges and spikes as well lightning strikes, which can contribute to the deceased life of a xfmr? Then there is harmonic content that can cause heating.
A 3ph xfmr installed without considering harmonic content can be subject to overheating with out being overloaded. A complex that was designed and used where harmonics was not an issue probably has a general-purpose xfmr (s). Why should the builder go through the expense of paying for K-factor xfmrs? The tenants move out and the area turned into an office space with lots of office space and the switching power supplies. Light fixtures are changed to the more current energy saving electronic ballasts.
Now you have a run of the mill general purpose transformer faced to deal with mega harmonics.

This may have been more than you ever wanted to know about transfomers.
Yes, determining xfmr life expectancy is like shooting at shadows.
 
For the IR remove the top front cover is then easily removed on most transformers. You can also shoot down and obtain some information also try shooting up if this is a raised floor you might be able to remove a tile.

Check the manufacturers literature forced air cooling in some situations will raise the KVA rating of the transformer. Forced air cooling whether allowing additional kva or not would still be a good idea.
 
Status
Not open for further replies.
Top