Temp Rise of Transformer

Measure it.

Its an addition.

When it was 20 deg C before and is now 25 deg C the rise is 5 deg C.

20 + 5 = 25

What do you really need to know?

See[for instance]:

http://ecee.colorado.edu/copec/book/slides/Ch15slides.pdf

sreerampkj:

Temperature rise calculation comes handy from

Heat Energy = Mass * Specific Heat Constant * Temperature

Temperature is reduced by cooling effect due to heat losses else it will be exactly what you calculate from above formula.

Sree Ram,

First you shall know the total losses (Core and Cu losses).

Then the mass of Cu, Steel and insulating oil and tank and their Specific Heats.

From above calculate the individual rise of temperature of Winding and core.

You will find the rise of core may be lower than winding and hence core along with oil and tank will work as heat sink for the complete mass, absorbing excess heat of Cu.

This will give you average temperature of transformer.

As a thumb rule take 10°K as difference between hot spot and average temperature.

Since this is a dynamic cycle of heating and cooling, heat is being generated in the winding and core and is being dissipated from surface of tank.

From surface area of tank and emissivity of steel (material of tank) and Δt (above average over ambient) you can calculate the heat loss from the surface.

Your aim shall be to achieve average temperature of 10°C below working temperature of insulation. For example for paper of coils is Class A or Y with limit of 70°C or 90°C (I hope that my memory is correct for working temperature for class of insulations). That means an average of 60°C or 80°C.

Remember 10°K difference is for hot spot temperature.

If dissipated heat is not available to achieve above, depending of rating of transformer, increase either surface area by adding tubular of flat tubes+fins, forced air cooling of forced air force oil cooling (ONAN or ONFA or OFAF).

If you are designing a very small transformer of few VA up to 2500VA then, commercially used current density will ensure operating temperatures within limit.

For small dry transformers now a days supper enamelled conductors up to Class H insulation (175°C operating temperature) are available.

If not sure, can manufacture first prototype with class H insulated conductor. Measure surface temperature with instrument (not mercury thermometer - it will be false as Hg is conductor and will heat up due to eddy current induced by leakage flux of winding), Disconnect from supply, start a stop watch, after noted time measure resistance of each winding, plot a curve of resistance vs time and manipulate winding temperature at switch off (Zero time).

From initial resistance and increase in resistance calculate temperature rise (Rt = Ro *(1 + Alpha * Δt)). This will be average temperature of winding. Add 10° to find hot spot temperature. This shall be less than specified for class of insulation. If it is too less, you can go for Class F or B depending upon what you achieve. Or else can increase the current density.

Above procedure for smaller transformer can also be applied to test large size prototype transformer. In case you are designing a large size transformer - testing engineer in the factory must be aware how to test for temperature rise.