EE Response
EE Response
I saw a comment "I'm not an engineer, but what I'd do is..."
So here's a response from an engineer....
... which is no more qualified than a response from an electrician, because we're all following the same code! specifically NEC Table 450.3(B)
In addition to adherence to the code, there's the other major factor... the load.
If the load has significant inrush potential, size the primary OverCurrent Protective Device (OCPD) to 250% with cable to match. Secondary at 125%. That's also the safest cover-your-a$$ call. Any loads that trip the primary OCPD are too large for the transformer. It's also the most expensive approach for the user because the wire is so much larger.
If the load is light, resistive in startup nature, or there are many many small loads, then you could reduce the size of the primary OCPD. Even to the point where you could use 125% primary only protection per the table.
But the risk there is the magnetizing inrush of the transformer. The primary OCPD must be able to withstand that inrush, which can be 8-12X the fla of the transformer, for an 8-20 cycle duration...
The last engineering firm I worked for standardized on 175% for the primary. It hasn't failed them in the last 5 years.
Bit it's hard for me to drop down from the 250%, especially now because I'm in the critical power market, where my mission is to maximize uptime.
The best way to look at this is on a Time-Current-Curve where you can plot the characteristics of all the devices in order to assure the transformer magnetizing inrush is tolerable with the chosen OCPDs.
John M