Compare old to new transformer losses

[dithomas]

I am searching for a way to determine the losses on 40 year old Federal Pacific 1000kVA 13.8kV - 480 delta delta transformers in order to compare them to a new modern transformer. A saving in operating cost could justify replacing the 14 FPE transformers with new Delta Wye units.

There are 14 transformers in groups on various 13.8KV circuit. An example is one circuit with four transformers on the same 13.8kV feeder. there is no metering on each transformer primary only the Eaton IQ4000 at the 13.8kV breaker. The load on each transformers is approximately a 500 amp steady load. This has been the loading for the entire life of the transformers.

Is there a way to determine the losses on one transformer?

The four single ended unit substations have tie breakers and a bus tie between them and can be paralleled to share load and one if the four transformers can be unloaded.

Any ideas? Thank you in advance.

 

[DTR2011]

Carefully evaluate the effects of Xfmr efficiency vs. potential Arc-Flash problems.

 

[mgtrp]

With regards to the original question, for the purposes of evaluating your losses the best place is to take them from the transformer test report although I doubt that you'd be asking the question if you had those. Otherwise, you're into measuring them yourselves on one of the transformers, or doing an estimate based on prevailing technologies at the time. Reports such as this may be of some use if you want to try to estimate.

And to DTR2011 - To satisfy the curiosity of someone who doesn't have a lot to do with arc flash studies, what is the effect of a more efficient transformer on arc flash?

 

[jghrist]

Measurement of the losses seems the only accurate way. Measurement of input kW vs output kW while under load would not be very accurate even if you had meters on the primary. Load and no-load losses are probably less than 1%, which makes CT and meter errors very significant. Don't forget no-load losses, these are generally more valuable on a per kW basis than load losses because they are there all of the time.

I really can't imagine that the loss savings would pay for replacing the transformer over whatever the remaining life of the old transformer is.

 

[ScottyUK]

The theory behind the standard open-circuit and short-circuit tests are standard fare for an undergraduate university power engineering course and covered in numerous texts.

If you are determined to do this the hard way and actually measure the losses then a rental generator makes a good and relatively cheap high current injection source. You'll need to substitute a small DC power supply in lieu of the AVR so you can directly control the field in order to vary the voltage from zero up to rated, and the set needs to be rated for whatever current you need. Don't undersize the shorting bars.

 

[dithomas]

jghrist,

I agree with you that there would not be much of a pay back based on losses. The issue is what is the remaining life of the old transformers and how does one justify replacement prior to failure that will cause production down time.

These plant have no trouble upgrading production equipment to the latest greatest and power them with 45 year old ungrounded transformers!!!!

I suppose there is some justification for eliminating the ungrounded secondary for a solidly grounded WYE. They have experienced drive and production computer equipment issues.

Dan

 

[davidbeach]

Look for the original transformer test report. The impedance test sheet should include loss information.

 

[mgtrp]

Bear in mind that you are not evaluating the benefits of the improved efficiency against the total cost of replacing the transformers, only the additional cost of replacing the transformer now versus replacing it in a few years time, or in the event of failure (this is a net present value type analysis). The savings from improved efficiency can form a fairly significant portion of the economic case to replace an old transformer when evaluated on this basis.

 

[dpc]

I suppose there is some justification for eliminating the ungrounded secondary for a solidly grounded WYE.

The justification has been well understand since the 1960s. Since you have an ungrounded 480 V system now, conversion to a high-resistance ground 480 V system would change virtually nothing in terms of operations, yet greatly help to reduce voltage transients. This will in turn, reduce motor insulation failures. I'd recommend looking into to that option.

As to your original question on losses, I think the best option is to try to find typical losses for transformers of that size and vintage and then compare to new units. Keep in mind that efficiency is something you pay for - manufacturers can improve efficiency by various methods, if you are willing to pay the price.

But I don't think increased efficiency is going to give you a lot of ammunition for replacement. A 40 year old 1000 kVA transformer is at end-of-life and I'm sure it is fully depreciated. I'd focus on:

Cost of downtime in the event of a transformer failure (it's when not if)
Reduced equipment insulation failures by converting from ungrounded delta to some type of grounded wye
Safety