Transformer losses

[ronmath]

Does anybody know the way to calculate the savings in transformer losses between an existing 150 KVA transformer and replacing it with a better sized 30 kva transformer?

Thanks.

 

[gar]

110727-2041 EDT

As an approximation break the losses into two categories --- core losses and I^2*R losses.

If the design criteria and core material are similar for the two transformers, then the 30 KVA has about 1/5 the core loss of the 150 KVA. This is fairly simple to determine. Measure the unloaded input power to each of the transformers.

The resistive losses are more difficult to estimate. Likely the load is not a constant, therefore it is time varying. But consider a 30 KVA load on the 30 KVA transformer, basically full load. This same load on the 150 KVA transformer will produce much less resistive loss. See if you can determine how to estimate this.

Thus, you need to consider the relation of the core losses to the resistive losses. If the smaller transformer is not generally loaded very much, but is required for peak load, then it might be the better choice.

I am not where I can look up any typical relationships of core to resistive losses.

.

 

[zog]

Core losses are very minor with modern transformers, depending on quality (core materials and laminations used to mitigate hysterisis and eddy current losses) and can be less than 1% of design rating, so assume 1.5kVA compared to 0.3kVA, thats (less than) 1.2kVA, or around 1kW or 24 kwh per day (Around $1 per day, and that is pushing it). How long will it take to pay for a new, smaller transformer? (Don't forget to factor installation time and material costs, and down time)

 

[ronmath]

Zog,

At 24KWH per day, with an average of $0.10 per KWH that would be $2.40 per day times 365 days per year, or $876 per year. That seems like it would be a decent payback over the life of a transformer. Am I missing something here? I haven't taken the resistive losses as Gar has suggested, but the NEC calculated load for the space is less than 30 KW, so actual demand load would likely be 1/2 of that.

 

[ronmath]

View attachment 20110728080008953.pdf I went to Square D and in their lean tools calculators they have a nice savings calculator for transformers (see attached results). The difference between the 150 KVA and 30 KVA transformers with similar loads was 810.44 KWH or about $81 per year, not enough to replace on it's own. However it showed if you have a standard 150 KVA and went to an EE 30 KVA, the savings would be around 8071 KWH per year or about $807.10 per year savings. A shorter payback worth considering.

Thanks for your help

 

[zog]

Zog,

At 24KWH per day, with an average of $0.10 per KWH that would be $2.40 per day times 365 days per year, or $876 per year. That seems like it would be a decent payback over the life of a transformer. Am I missing something here? I haven't taken the resistive losses as Gar has suggested, but the NEC calculated load for the space is less than 30 KW, so actual demand load would likely be 1/2 of that.

Well I used 1% as the basis, when in reality it is less than that, and $0.10/kWH is a residential rate, I was assuming this is industrial where most facilities are about half of that. Your $81/year sounds about right, but as I mentioned it depends on the type of transformer.

 

[mull982]

why are we not considering the I^2R loss calculations when determining cost savings? It seems like we are only focusing on core losses.

 

[zog]

why are we not considering the I^2R loss calculations when determining cost savings? It seems like we are only focusing on core losses.

Because we don't knwo the load, and it would be the same regardless of size so pretty much cancels out.

 

[mull982]

Because we don't knwo the load, and it would be the same regardless of size so pretty much cancels out.

Is that really the case?

Lets say that both transformers were 5% impedance transformers. wouldn't we have to convert them to an equal base to compare apples to apples with resistance losses.

So a 5% 30kVA transformer on an 150kVA base would be 15%. So wouldn't we be comparing the 5% impedance on the 150kVA transformer vs the 15% impedance of the 30KVA unit on the same base? In other words wouldn't the smaller transformer have a higher imepdance for a given load?

 

[zog]

In other words wouldn't the smaller transformer have a higher imepdance for a given load?

Yes, but we still don't know what the load is.

 

[mull982]

Yes, but we still don't know what the load is.

O.k. but we are agreeing that for the same load on both transformers the I^2R losses on the 30KVA transformer due to the fact it has a higher impedance, and in order to compare these losses you would need to converter both transformers to the same base?

 

[rcwilson]

O.k. but we are agreeing that for the same load on both transformers the I^2R losses on the 30KVA transformer due to the fact it has a higher impedance, and in order to compare these losses you would need to converter both transformers to the same base?

I think there?s confusion between impedance and losses. Transformer impedance is largely inductive with a small resistive component that represents the load loss. Nothing in the impedance represents the excitation losses, the no-load losses.

In this situation, we don't care about % or per unit impedance or losses. We want to compare the total losses in watts of the two transformers under the same loading conditions. To do that we will need the no-load loss and load loss data for each transformer and an estimate of the actual load profile.

From an old Eaton Consulting Engineering guide:

150 kVA, 115C rise, DT-3 dry type has 700 W no load and 4430 Watts load losses.

The 30 kVA has 250 W no load and 1460 W load losses.

Assume actual load is 25 kVA continuous, or 83% of the 30 KVA and 2.8% of the 150 kVA.

Load loss varies with the square of load:

30 KVA has (83%)^2 x 1460 W = 1014W loss,
150 kVA has (2.8%)^2 x 4430W = 123 W loss.

Add the no load loss:

30 kVA = 1014 + 250 = 1264 W = 1.264 kW
150 kVA = 123 +700 = 823 W= 0.823 kW,

The 150 kVA uses .44 kW less at 25 kVA load.

Drop the load to 17.76 kVA and the two units have equal losses.
At 10 kVA load the smaller transformer saves 0.3 kW.

Conclusion - you need to know the load profile and the actual loss data of the existing and proposed transformers. That data is not on a nameplate.