Minimum AWG wire that can feed a sub panel

[Adamjamma]

Although rarely enforced here, I am of the opinion the 210.119C)(4) along with 210.70(A)(2)(2) requires a minimum of a MWBC.

If I remember correctly, though it has been over 20 years since I last did any USA Electrical, garages are only required a single 20 amp circuit, not an MWBC.
Nor requirement for subpanel.
Now, if you are putting a sub panel in, even here in Jamaica, you size the feeders to your main breaker, wither it is an actual main breaker or a side feed or reverse feed breaker, that has to be locked down so it can not be pulled by accident.
I personally would use 6 gauge wire and a 50-60 amp breaker, then would have plenty of capacity for running saws, air compressors etc- and personally would run it as 240/120.. Plus would probably use a 2 inch conduit from house to it, so could pull bigger wire if needed later..
But it is always customer checkbook against job requirements.. NEC is minimum code, not required design maximums.

 

[suemarkp]

Code now requires a 20A circuit to serve receptacle outlets at each vehicle space and can't have any other outlets except in that garage. This implies that lighting in the garage has to be handled by a separate circuit. So 2 circuits are needed for a garage.

 

[don_resqcapt19]

They were very close to fixing all of this in the 2020 code. The first draft made the EGC table work like the grounding electrode conductor table...that is the size of the EGC would be based on the size of the ungrounded circuit conductor. They miss a lot of details in the first draft that made the change unworkable and did not have enough time to fix it for the second draft. Not sure why it was not looked at again for the 2023.

 

[david]

They were very close to fixing all of this in the 2020 code. The first draft made the EGC table work like the grounding electrode conductor table...that is the size of the EGC would be based on the size of the ungrounded circuit conductor. They miss a lot of details in the first draft that made the change unworkable and did not have enough time to fix it for the second draft. Not sure why it was not looked at again for the 2023.

Interesting, why not just leave the current table alone. And just require a full size equipment ground if the ungrounded conductors are increased in size.

Or leave the current table alone and require a voltage drop calculation on the eguipment ground using the overcurrent ampacity, when the ungrounded conductors are increased in size. (Not sure of the voltage drop calculation having the end result needed)

They should do something to simplify this.

 

[don_resqcapt19]

Interesting, why not just leave the current table alone. And just require a full size equipment ground if the ungrounded conductors are increased in size.

Or leave the current table alone and require a voltage drop calculation on the eguipment ground using the overcurrent ampacity, when the ungrounded conductors are increased in size. (Not sure of the voltage drop calculation having the end result needed)

They should do something to simplify this.

In my opinion the very simplest would be having the size of the EGC directly based on the size of the ungrounded conductor. That would result in no changes for most installations and eliminate the discussion that results from 250.122(B) as the ungrounded conductor size would directly determine the EGC size.
I would even be fine with using a table like that for the individual conduits or cables of a parallel installation, that is the EGC size based on the ungrounded conductor size in the raceway or cable. That would eliminate 250.122(F).
This sizing method has been used for the size of the grounding electrode conductors, and supply side bonding jumpers for a long time.