Mixed brand MCBs and EICR coding

[SimonH2]

Applying the highest standards of professional responsibility and duty of care, do you KNOW that they fit properly and do you KNOW that there CANNOT EVER BE ANY problems from mixing?

Playing devils advocate, replace "problems from mixing" with "problems using all the manufacturer specified parts" and answer the question. In truth, we know that the answer to both questions is the same - because we know from empirical evidence that there are occasional faults even when all the correct manufacturer supplied parts are used. Thus either way, the answer is "no" - either way we cannot KNOW that there CANNOT EVER BE ANY problems.

A more interesting one is, what if you keep (say) the outer box and replace all the breakers - just take all components out but the main switch and throw a load of RCBOs on the DIN rail left. The argument about disturbance between adjacent MCBs or RCDs from different manufacturers cannot apply because they are all the same ones that would be side by side if we used the manufacturer's box. If we replace the main switch (and bus bar) as well, then there's no argument about differences in position of the terminals.
As it is, I considered that as an option for our house - it's got an old metal cased Crabtree CU, and just swapping the internals would be quite easy. And I think the case and DIN rail is probably "at least" as good a quality as you'd get on a new unit. But as I want to add circuits, it would still be notifiable so very little saving and I'm not happy with it's location.

 

[stillp]

However, if all one has is a theoretical argument as to the mechanism whereby a certain set of circumstances could/might result in some danger, then one needs data in order to quantify the risk of that 'potential danger'.

Perhaps one does, but one need not "quantify the risk", even if such a process were possible from historical data. One need only recognise that there is a risk, which can be easily and completely averted by not mixing components of a CU that come from different manufacturers.

if they could show me empirical evidence

Why on earth would they do that? I don't imagine that "we must ask JohnW2 what he thinks" is a part of most manufacturers' planning process.

However, although I may be wrong, I suspect it's not like that. I suspect that what they know is a theoretical mechanism whereby mixed-make devices could/might result in a danger, without ever having demonstrated that it actually happens (at all, let alone consistently) in practice.

Your supposition that you might be wrong is correct. It is known that MCBs and RCDs are susceptible to (small) changes in performance due to external influences, including temperature and magnetic fields. That is not a theoretical mechanism, but can be demonstrated in laboratory tests. Manufacturers cannot test their products in every possible combination of installations with every other manufacturers' products, so all they can do is insist that their products are used as instructed. As I believe I've mentioned previously, I am aware of the existence of a test report that identifies a significant effect on protective devices that arose from use alongside devices from another manufacturer. That report is confidential to the company that commissioned it, though I have their permission to use it as evidence in a legal issue.

I really cannot discuss this topic very intelligently.

I fully agree.

 

[kai]

A more interesting one is, what if you keep (say) the outer box and replace all the breakers - just take all components out but the main switch and throw a load of RCBOs on the DIN rail left. The argument about disturbance between adjacent MCBs or RCDs from different manufacturers cannot apply because they are all the same ones that would be side by side if we used the manufacturer's box. If we replace the main switch (and bus bar) as well, then there's no argument about differences in position of the terminals.
As it is, I considered that as an option for our house - it's got an old metal cased Crabtree CU, and just swapping the internals would be quite easy. And I think the case and DIN rail is probably "at least" as good a quality as you'd get on a new unit. But as I want to add circuits, it would still be notifiable so very little saving and I'm not happy with it's location.

This sounds like a great idea to upgrade an old consumer unit, where you can't source parts anymore. The main switch can probably stay, as long it is a cable in and cable out device, with no direct attachment to either busbar in the CU, in the case of a dual rcd only unit, as the output of the MS is two pairs of cables that supply the RCD's in turn. Saves a lot of red tape! You could upgrade an Old Wylex NN board to modern standards this way (if it has two neutral bars in it).

 

[stillp]

You could, but it would not comply with BS EN 61439-3 and hence BS7671.

 

[EFLImpudence]

As I indicated earlier and using only logic:

If ALL CU components must comply with BS EN 61439-3 yet some CU components cannot be mixed (assuming physical dimensions are the same) with other CU components because of "type testing" (nearly as bad a phrase as like for like) then some of theses components must be superior to others; these superior ones must therefore exceed BS EN 61439 and so are better than the ones which only just comply.


It therefore follows that ALL CU components comply with BS EN 61439-3 so the superior ones must be able to be mixed with the inferior ones but not vice versa even though the result would be identical.

 

[flameport]

I don't see why most of these wrong MCBs are ever fitted - it's not as if MCBs of the majority of makes are difficult to obtain or cost vast amounts of money either.
If installing a new circuit or whatever, surely those installing it must have already seen the CU and noted the manufacturer of it?

Here's one which is less easy to see:

The offending item:

 

[stillp]

If ALL CU components must comply with BS EN 61439-3 yet some CU components cannot be mixed (assuming physical dimensions are the same) with other CU components because of "type testing" (nearly as bad a phrase as like for like) then some of theses components must be superior to others; these superior ones must therefore exceed BS EN 61439 and so are better than the ones which only just comply.

Eh? I don't follow that. Type tests (a term used in most product standards) are to verify that the product meets the requirements of the standard. Type testing MCBs will verify e.g. the trip current, temperature rise, etc are within limits. Type testing a CU will include verifying that the MCBs function as specified, and don't interact in a way that makes them fail to conform to the specification, termianls don't overheat, materials meet the dielectric and flammability tests, etc.
Yes, some manufacturers products are better than others, but all manufacturers (except the counterfeiters) will type-test their CUs using their own components. No doubt some manufacturers' products will sail through with room to spare, others' will struggle to scrape through the tests, but even if you knew which were the better ones you would still not be able to fit them in another manufacturer's CU and still claim conformity.

 

[ban-all-sheds]

Of course I don't.

No, of course you don't.

You DON'T know, but you ask whether it is reasonable to consider whether problems could arise.

You DON'T know, but you suggest that it might not be appropriate to code it at all.

You DON'T know, but you (not for the first time) talk about using "common sense" as an alternative to regulatory requirements.

You DON'T know whether heterogenous products really do fit together properly and safely yet you posit a situation in which they are dimensionally identical and "fit together".

The question was "Would you code it on a commercial EICR and if so, what code and under what reg?" Reminding you once more that you would be (one would hope) applying the highest standards of professional responsibility and duty of care, and as you don't KNOW that they fit properly and you don't KNOW that there CANNOT EVER BE ANY problems from mixing, how can you ignore it, or just "make a note"?

You have to flag it as unsatisfactory, and needing further investigation.

 

[EFLImpudence]

Eh? I don't follow that.

OK.

Type tests (a term used in most product standards) are to verify that the product meets the requirements of the standard.

Yes, so they ALL must.

Type testing MCBs will verify e.g. the trip current, temperature rise, etc are within limits.

So, they are ALL compliant.

Type testing a CU will include verifying that the MCBs function as specified, and don't interact in a way that makes them fail to conform to the specification, termianls don't overheat, materials meet the dielectric and flammability tests, etc.

So, ALL manufacturers will have done this.

Yes, some manufacturers products are better than others, but all manufacturers (except the counterfeiters) will type-test their CUs using their own components.

To the same specifications.

No doubt some manufacturers' products will sail through with room to spare,

So, they are superior to the others.

others' will struggle to scrape through the tests,

That's emotive; if they got through then they comply.

but even if you knew which were the better ones you would still not be able to fit them in another manufacturer's CU and still claim conformity.

So, it's just a rule with no basis.

If we did know which were the superior ones would you say there is nothing against putting them in the inferior product apart from the rule?

 

[kai]

I will always consider a pic-n-mix consumer unit to be a bad thing, they are purposely designed these day with obsolescence built in, to make you invest in a whole new board a few years down the line. This is why breaker designs are changed to keep with current fashion, the cost of re-tooling everything is eclipsed by the new sales that it will generate by people needing to replace their boards.
The dangerous alternative is to buy second hand breakers off ebay, which may have interrupted a heavy short circuit just in excess of their breaking capacity (at 0.15 power factor!) a few times, so may be risky to buy.

 

[JohnW2]

Perhaps one does, but one need not "quantify the risk", even if such a process were possible from historical data. One need only recognise that there is a risk, which can be easily and completely averted by not mixing components of a CU that come from different manufacturers.

Life is full of risks, and only if they are quantified can one decide whether they warrant any 'avoiding action'. Even within the context of electrical installations, I'm sure I could think of many 'risks' which could be "easily [and maybe completely] avoided" by taking certain actions, but which have such a vanishingly low a probability of
occurrence that such 'avoiding action' really is not needed.

if they could show me empirical evidence

Why on earth would they do that? I don't imagine that "we must ask JohnW2 what he thinks" is a part of most manufacturers' planning process.

You're twisting my words (even after correctly quoting them!) again. There is no reason on earth why they should ask me anything, and I certainly haven't suggested that they should. I am being told of these 'known potential dangers', so I would like to see the nature of the evidence and, ideally, given some indication of the probability of these 'potential dangers' becoming realised dangers in practice.

It is known that MCBs and RCDs are susceptible to (small) changes in performance due to external influences, including temperature and magnetic fields. That is not a theoretical mechanism, but can be demonstrated in laboratory tests.

I would think it inevitable that influences such as temperature and magnetic fields can result in "(small) changes in performance" of MCBs and RCDs. What I don't have a clue about are the magnitudes of these "small changes in performance" and the extent to which external magnetic fields and temperature changes due to MCBs and RCDs vary between manufacturer. I am being asked to have 'blind faith' in something without any of the evidence being available to me - and that goes totally against the most basic principles of my scientific upbringing.

I also wonder how these "(small) changes in performance", even if it can be demonstrated that such can be brought about by the proximity of a different-make device, translate into "potential dangers". I wonder, for example, how these "(small) changes in performance" due to the proximity of different-make device compare with, say, the changes in performance that can result from a 20°C change in ambient temperature.

Manufacturers cannot test their products in every possible combination of installations with every other manufacturers' products ...

Of course they can't, but it would still be 'nice' if we could be told of just one laboratory test which had revealed a "potential danger" resulting from the co-existence of devices from any two manufacturers.

As I believe I've mentioned previously, I am aware of the existence of a test report that identifies a significant effect on protective devices that arose from use alongside devices from another manufacturer. That report is confidential to the company that commissioned it, though I have their permission to use it as evidence in a legal issue.

It's all very well for you to keep telling me of the existence of 'facts' supporting your viewpoint which (even if for good reason) cannot be made available to me - but, as above, accepting things on the basis of 'blind faith' is not something that I am used to doing!

Kind Regards, John

 

[JohnW2]

The question was "Would you code it on a commercial EICR and if so, what code and under what reg?"

That was actually the second of two questions, the first being more relevant to a DIY forum. However, if the "commercial EICR" relates to a DB which is not "under the control of ordinary persons", how would you code it on an EICR and, in particular, which regulation would you cite as having been contravened?

... and, perhaps more to the point, if, as we are being told, this regulation exists because of a 'known potential danger' of mixing devices of different manufacturer in the same DB, why does this 'known potential danger' suddenly cease to be of concern if the DB is "under the control of" someone who is not and "ordinary person"?

Kind Regards, John

 

[stillp]

If we did know which were the superior ones would you say there is nothing against putting them in the inferior product apart from the rule?

Which rule? The rule that they must be type-tested? Unless they've been tested in the different manufacturer's enclosure, then they haven't been type-tested. It's quite likely that they will pass the tests in one manufacturer's enclosure and not in another. That's the whole point of type-testing the complete assembly.

 

[JohnW2]

It's quite likely that they will pass the tests in one manufacturer's enclosure and not in another.

"Quite likely"? Do you seriously believe that??

Kind Regards, John

 

[kai]

If ban could have his way, we would go back to the Wylex Standard Range system (Plus RCD's), where everything is locked down. No pick-n-mic possible if that were the case, and secure terminals too, with those twin screw terminals. Do you bit Ban, as you have such an influence on those who write BS7671, you could make a safer Eighteenth edition come true.