121108-1906 EDT
My guess is that your 12x factor is probably the least to use unless you have a special breaker designed to avoid this problem. With new high efficiency transformers the problem may be worse than with older transformers. This will be very much a function of the trip-time characteristics of the breaker and its instantaneous response.
With different loads and phase angles of the loads at power turn-off there will be different values of residual flux. Thus, all transformers may not be in the same state as indicated by kingpb. Heavy resistive load at turn-off should produce a lower residual than an unloaded transformer.
See my photo P6 at
http://beta-a2.com/EE-photos.html . This is from standard transformer iron of a transformer using general industry standards for design of about 25 years ago.
I am not at all close to transformer or breaker design today so my comments are broad and general. Some long time ago I designed a breaker concept to circumvent the problem of inrush current. This is described in patent number 3,299,322 issued Jan. 17, 1967 and initially filed July 23, 1962.
A broad claim read as follows:
" 1. In an electrical circuit comprising a source of electrical energy and a load, an electronic circuit protector for interrupting the electrical circuit upon the occurrence of a predetermined condition therein, said electronic circuit protector comprising a signal responsive switch electrically connected to the source and load, means for sensing electrical conditions in the circuit having an output signal, a switch control circuit normally responsive to the output from said means for controlling the condition of said switch, an inhibiting circuit for inhibiting the output of said sensing means for a predetermined initial period subsequent to turn-on of said electronic circuit protector, and a desensitizing circuit for desensitizing said switch control circuit for a predetermined period of time subsequent to the inhibited period."
The intent of this circuit and patent was to overcome the inrush problem and yet provide very tight overload detection after the initial inrush. We could sense a 1% overload after the large inrush current had passed.
Do any of the electronic breakers perform this function today? I do not know, but they should.
.