EMC

There is not a serious outcome unless the steel is fairy thick, certainly much more than the 1.5 to 2mm of domestic CUs. The problem is really overstated with modern enclosures, but comes from long ago when switchgear had thick cast iron casings, often 3/8 to 1/2 inch thick. Significant heating is then possible with high current cables, but the circuit reactance is minimal with thin metal. Thick cast iron has much higher losses than thin steel, but changing this suggestion of a problem has never happened, and may still be seen with some heavy duty switchgear. At 60A or even 100A there is not much problem, and as in all heating this is a long term sustained current, not a short peak. Under fault conditions at several kA there is a significant force on cables which are not closely balanced or are through separate holes, but in a domestic this is unlikely to cause any damage. In a large installation with much higher PSSC the force can be very large, sufficient to cause significant physical damage to casings and cables. It should probably be a C3 for domestics, not ideal but not dangerous.


I forgot to say that single conductors should not travel through steel containment as this acts as a long path through the material and losses can be quite high

There is also a lot of confusion out there about what is an eddy current loss - electric current and ohmic, and what is a magnetic hysteresis loss.

The problem with a unbalanced currents in conduit has two parts - magnetic, which would still be a problem if we sliced the tube into short cylinders, but would be removed it we slit the tube length-ways, and closed the gap with braze metal or even soft solder, and electric effects, where a voltage appears between the ends of the conduit, that is then shorted by other metalwork - often other conduits on the same boxes or plates, and where even if the tube was all brass we would still see the problem, and the solution is to cut the current path into more than once isolated cylinder.

(this is where folk can come unstuck with AWA singles and hot gland plates)


In a thin wall box, only the magnetic effects are seen, and unless the hole is a tight fit about the cable, and the cable is heavily loaded, the magnetic path is going to be long enough not to be present much additional loss. Making the hole bigger with a slot, (or to dogbone the holes for flow and return current)

or using a non-magnetic enclosure are the fixes. Again, slots can be filled with soft solder or braze, or at the cheap end car body filler or at the really cheap end, left open and covered with gaffer tape..


Even without any attempt to enlarge the magnetic path, a 100A tail through a 20mm hole typical galvanized back box will not get anything very warm, so in nearly all domestic with mainly 32A circuits and less  it will be a non-issue. The problem is the magnetic losses of steels are very variable, and the sort of stuff used for back boxes is not optimised for that so it is hard to put  safe upper limit, and folk like to be cautious.

(If you have an arc welder and some spare socket boxes you can try this for a bench demonstration wrapping the  tails through the box holes a couple of times to get N times the current - In my experience,  the only thing that gets far too hot is the welder itself and the cable in the normal ohmic sort of way - I have a cheap Chinese-made welder rated for about 10 minutes an hour at full power...)


It is up there with not going swimming after lunch as being far less of an issue than the old folk told you.

A JW practical demonstration for interest (summary: ~120A, ~2C rise in this example)

That is like a rather more business-like version of my welder demo. I'd have struggled to see 2 degrees rise in the metal with my set up. Based on comparisons of touching with fingers, you need to send the nominal  150A welder current  round at least twice through the hole to get something in feeling range,and then the cable is also getting warm as part of the normal run of things. Nice to have the results backed up though, and I stick with the rule of thumb that upto to 30A through a normal galvanized back box will do nothing worth worrying about.

I have an example of this, tri rated singles through a sheet steel enclosure, each in a stuffing gland through its own hole, not slotted.


its loaded at 400a 


the enclosure is about 1.8m tall by 600 wide


I can feel the whole cabinet is slightly warm, it always was, but there is a noticeable temperature rise around the cables in the steel, and the cables are warm where they are passing through the enclosure. Only a few degrees difference.


If I move one of the cables, plenty of slack, I can feel the 50Hz vibration


Its been like that for many years, I suspect the large sheet steel enclosure is a good heat sink