In the circumstances you describe, I think I would have no hesitation in swapping the existing breaker with a suitable alternative. I would, however, carefully set out the engineering judgement that I employed to validate that all reasonable precautions were taken to ensure the equipment and connected circuits remained safe. 

There would be plenty who would swap it out without a second thought. If you are overthinking things, it was good to be reminded about this significant issue.

I think that the concern is entirely valid and agree with lyledunn.

You mention your employer, so as an employee (as opposed to contractor or subbie) you have nothing to worry about. I would interpret, "person" in the last full line of Note 2 to 536.4.203 to be the legal person rather than the flesh and blood individual, i.e. the company which employs you.

(Good title, BTW.)

My employer thinks that a C100A MCB is a C100A MCB

Certainly not always. In a DB or CU there are obvious physical differences such as the height of the terminals from the DIN rail which can mean they don't match bus-bars etc. If your connections are simple wires however that's probably less of an issue. Likewise in a CU MCBs may be required to break more than their rated breaking capacity (e.g. 16kA rather than 6kA) and different manufacturers may have adopted different approaches to achieving that (anything from relying on the let-through of an upstream fuse, to the enclosure containing arc products). Again if you're no relying on an extended breaking capacity that's less of a worry. Then there are more subtle things like the way heat generated by one MCB can affect the tripping characteristics of its neighbours which may very from one make/model to another - but if your MCB is on its own or separated from its neighbours by a blank module or two, that can be mitigated, even if it was an issue to start with.

Yes, in principle, you become the manufacturer with all the responsibility that comes with that; but from a real world perspective in most of EN land people just assemble CUs and DBs on site with a complete mix of manufacturer's components and no-one blinks an eye - so (within limits) the actual risks would appear to be vanishingly small.

   - Andy.

Thanks Andy. Not relying on upstream protection, so a reasonable consequence of something going wrong would be an earlier OCPD operation due to it being warmer? 

Thanks for your response Chris. If you had to reference this engineering judgement to a British Standard to back up the argument what particular items within 61439 should I be concerned with?

Thanks Lyle.

A change in trip thresholds due to warmer neighbours is something that afflicts all breakers with a thermal part to some degree. It is a poorly thought out design where it actually matters however....  I assume your 100A breaker does not have a 99 amp 24hour 7 day load but if it does then giving it  a bit of fresh air is a good idea, whether the neighbours are the same make of breaker or another make ! Most makers publish curves or tables very similar to this (which is for MCBs but the principle is similar)

So for example at 55C most larger breakers behave more or less like the next size down....

The most dangerous consideration for mix and match is always the PSSC - is there an unlikely 'silver nail' worst case fault condition that could cause the replacement breaker to weld and fail to trip or to only disconnect dangerously late ?

If that happened, one has to ask what would 'let go' next ?- This is the use of an upstream 'death or glory' fuse of many hundreds of amps.  Generally if there is one in the supply somewhere then the safety case is a lot easier to justify.

(The alternative of the breaker casing falling apart and disconnecting never to reset again is a nuisance  but can be argued to be a perfectly safe and acceptable failure mode without the upstream fuse so long as all the burning bits that fly off are contained in the enclosure,..... this is the sort of thing we like manufacturers to test, so we don't have to.)

Mike.

E95 said:

If you had to reference this engineering judgement to a British Standard to back up the argument what particular items within 61439 should I be concerned with?

I am no expert in this matter, but BS EN IEC 61439-1:2021 requires that the design be verified. There are 11 different items in Article 10.1, but they include temperature rise and short-circuit withstand strength.

As with BS 7671, there is a provision which makes anybody who modifies the original design assume the responsibility of the original manufacturer.

The short-circuit withstand strength might be tested (destructively) by the original manufacturer, but verification by reference to existing components is permitted.

Perhaps there is an analogy with car tyres. Quite obviously, they need to be the same size as the originals. Then they need a speed rating which is at least as high. However, they can be made by a different manufacturer.

I certainly do not think that a panel should be replaced in its entirety just because one component has failed. That would be wasteful, but it does need sound engineering judgement.