Why do our electrical utilities use transformers way over their rated KVA?

Question

My parents' house and my house have different electric utilities, but both are rural electric cooperatives. We both have 120/240V (60 Hz) split-single phase service (this is in the United States state of Tennessee). My service is sized 600A. My dad's service is currently sized 200A, but it's getting upgraded to 400A because his property has outgrown 200A.

My understanding is that transformer sizing should be fairly simple. Volts times amps equals volt-amps. Size your transformer for the volt-amps. I also get that utilities tend to under-rate things and run them "hot," but this just seems extreme.

(Related question: I was never clear whether the "volts" in this equation should be 120 or 240. Is the transformer rating based on the combination phase voltage, or the split phase voltage?)

My transformer has "37.5 kVA" stamped on the side of it. 600A x 120V = 72 kVA. My transformer is 52% the size it should be. Not terrible, but still seems very under-sized.

My dad's transformer (before this morning) had 1.5 kVA stamped on the side of it. They "upgraded" it to prepare for his 400A service upgrade. Now it says 2.5 kVA. 200A x 120V = 24 kVA. 400A x 120V = 48 kVA. His old transformer was 6% the size it should be. Now his new transformer is 5% the size it should be. That's even worse, by a huge amount. That's just ... crazy small.

This white paper backs up my calculations. What am I missing here? This can't possibly be correct.

Answer 1

Distribution transformers have large masses of metal and (often) oil in play, they can run at large overloads for a LONG time before the heat builds to a problematic level.

Because of this they are often sized for the RMS load integrated over a period of several hours, which allows then to be far smaller then you would expect from a maximum rated breaker calculation.

Remember also that IIRC the NEC specifies that loads should not be more then 80% of breaker rating, so that 600A service is really 480A design maximum, and how often are you maxing that out for more then a few minutes at a time?

Incidentally if the breaker is 600A in each leg of a split phase service then the power is 600A * 240V = 144kW, but I would not be at all surprised to see a transformer somewhat smaller then that used, as I say integration times are a thing in this game, 50% overload during peak hours is a pretty standard place to be, but the load mostly disappears over night so everything gets to cool.

I have seen an amusing problem when a potter took on a light industrial building and installed two rather large three phase kilns (Only one could be run at a time, but once the first one starts cooling, spark up the second one), seems the local transformer did not appreciate back to back 12 hour raku firings.

I am somewhat surprised that 600A single (split) phase service is even a thing, over here if I wanted 144kW service it would be three phase.

Answer 2

Rural distribution wiring may have longer path lengths than urban, and my Rule of thumb may differ from your local codes, https://www.nfpa.org/-/media/Files/Code-or-topic-fact-sheets/ReferencedStandardsFactSheet.ashx .

Distribution line drop must be < 5% for rated current, thus the wire gauge must be reduced from AWG 8 to 5 for double current rating, but the demand kVA may not necessary be double for economical transformer choices if shared.

The transformer impedance of ~8%=Zpu of V/A rated(=V^2/VA_rated + I^2R of linear drop determines the load regulation of 5%. (My Rule of Thumb). They would adjust Xfmr-taps to keep Vnom +/-10% based on expected useage.

Typical utilities in N . America use 5% tolerance for generation and 5% for distribution to achieve Vnom to service.

Your father’s kVA rating appears to be low and incorrect. Estimate the size and weight for kVA/kg standards for a sanity check. Perhaps all they need is a power study and tap change to adjust for more line loss to support the load or else a costly extra AWG8 parallel feed and Xfmr upgrade.