db/cu and 521.5.1 Ferromagnetic enclosures: electromagnetic effects

Unless you are throwing hundreds of amps down the CPC, in normal operation,  or the enclosure is a very odd very thick super lossy steel, technically this is a non issue.  you could put a stud-bolt and a pair of wingnuts on each side, such that you had a wire on either side of the box wall, and the bolt carrying the current was itself made of steel, and in intimate contact with the box, and still see a temperature rise dominated by the normal I2R considerations.

There are a few things in the regs that charitably suggest that a blanket rule has been made when the case it applies to is infrequent, or less charitably, the level of understanding of magnetics among the writers of some technical parts is not good. 

From a high frequency fast transient point of view of wiring to an SPD you want to keep the loop area small, so running the CPC beside the L and N it relates to is sensible, but threading a few washers on the wire - equivalent to magnetic-only effects of the wall of a box is not going to add enough inductance to be measurable, as the ‘mu’ of the steel may be high at 50Hz, but it falls smartly at increasing frequency, due to the time it takes to turn a large no of spins around en-masse.  Even if this was not true, saturation effects mean that much like the RCD blinding, with high current  events, the core more or less is not present.

Mike

Unfortunately, there are many areas where the regulations make silly requirements due to old wives' tales. This is a primary one and as I have explained along with Mike and others the effect of sheet steel boxes is essentially zero. Large cast-iron cases are a different matter altogether, the characteristic being the length of the conductor inside the magnetic material. You can verify this yourself with a length of steel conduit and an arc welder, passing a large current through a single conductor inside the conduit will heat it significantly. Zoomup's picture failed to heat because the length is very small. Essentially this is a badly written regulation, and probably is copied from some IEC document, but getting technical editing of the regulations is very difficult. I did manage to get one completely faulty new regulation dropped, but it took a long technical paper, experimental proof, and a long time. Objections of regulations in the DPC are considered, but it takes a lot of comments to get a change. Keep up the good work!

In order to satisfy the naysayers, I have just measured the added inductance from 4 mm of steel washers threaded wires on a length of wire, it added 2 nH (Two nano Henries) to the inductance of the plain wire! This shows that the magnetic coupling is actually quite small, and at 50Hz the loss is only partly connected to the added inductance, it depends on the resistance of the coupled material too. As a comparison, the inductance of a metre of wire measured as a big loop measured around 1.642 uH, varying somewhat with the exact loop size and position as expected. The measurement frequency for this is much higher than 50Hz, so it may be that the material (steel has poor high-frequency magnetic properties) but measuring such small values at 50Hz is very difficult indeed. Even if the washers add a few micro-Henries at 50 Hz, the effect will be tiny. The poor high-frequency performance of Iron is useful for induction hobs and heaters and is widely used for manufacturing forged parts. Here though the length of the material is large and the coupling is a multi-turn coil, about 6 inches diameter in my induction hob. It heats a vertical wire very little but steel pans with 2 kW. 

Thanks for the tip about the demo using the buzz box welder and the conduit - it is pretty obvious after it has been mentioned, but I had not thought of it as a way to show the effect and explain why wires in steel conduit ought to balance flow and return.  That's this afternoon's activity  sorted then.  I now suggest that that one could also do an all cores in parallel version of the welder test with some steel wire armoured cable - like the conduit there should be a discernable effect, but the armour being wires with some in contact and some gaps, rather than a solid magnetic cylinder, confuses matters, if anything making the magnetic coupling effects weaker.

It is very telling that on the welder  the two wires come out of two separate holes in the steel plate, and while the front panel, well the whole thing actually, does look rather like a bomb has hit, there is no evidence of any extra heating around the wire exits.

Mike..

Thanks every one …  the knowledge from experience offered on this forum is much appreciated.

Nothing like an experiment, but it seems from even basic considerations and without experiment, the protective conductor requirement of being through the same aperture is a bit of an unexplained inclusion and a non-issue - at least outside of the extraordinary.

If current is significantly unbalanced (due to fault) that enough leaves via a protective conductor through a different hole (and possibly via other conductors too) to the line and neutral, then it isn't an issue.  It seems not much an issue for the line and neutral being in different apertures at least at 100A levels, but i was more intrigued by the protective conductor (and the different materials it might be) and why it was mandated in the Reg.  to pass through with the L&N.  Perhaps the writers will at some point explain further the reasoning they had.

In summary and with regard the Reg, not having the cpc/mec routed through the aperture with the circuit L&N in a steel db/cu, is currently a [foisted?] non-compliance, but there seems no real issue if it is done.

Thanks again

Good, that is that sorted. Are we all agreed that a non-compliance is C3 rather than C2?

So now what about a MPBC? As far as I can see, that is allowed to go through a hole of its own.

Chris Pearson: 
 

Good, that is that sorted. Are we all agreed that a non-compliance is C3 rather than C2?

So now what about a MPBC? As far as I can see, that is allowed to go through a hole of its own.

No, a non-compliance is a non-conformity that would give rise to danger. In this post my interpretation would be that the situation would be a non-conformity only. Personally I would not mention on any report.