BS EN 61439-3 - what it actually says vs what people think it says

[ericmark]

Only in collage and university libraries, the county library was a dead loss.

 

[SimonH2]

Thank you Simon, it has been a problem since I started as an electrician, that near impossible without spending a lot of money on the books to find out myths from fact

Indeed. As you might have guessed, I've got access to BSs online I don't need them for the day job , but it doesn't cost any extra for me to access them.

which is not helped by trade articles which are clearly designed to sell their goods not give facts.

And that includes articles where you'd expect the authors to know better As I mentioned, one article I came across was an IET Wiring Matters which you'd think ought to be "impartial" in such things.
As an aside, I used to subscribe to Which? One of the factors in cancelling (apart from the rising cost) was when they'd have reports on a subject area I know a bit about - and I'd be left thinking WTF at the errors/omissions/general inaccuracies

 

[JohnW2]

As an aside, I used to subscribe to Which? One of the factors in cancelling (apart from the rising cost) was when they'd have reports on a subject area I know a bit about - and I'd be left thinking WTF at the errors/omissions/general inaccuracies

I had almost exactly the same experience and also cancelled my subscription, at least partially for that reason (as well, as you say, rising cost).

Kind Regards, John

 

[mattylad]

Yes I have had access to BSOL via my local library web portal.
I had another search and have found the login page - http://www3.lancashire.gov.uk/onrl/BSI/BSiLogon.asp
Just need to find my library card number now.

Found it - but the only 61439-3 is from 2012.
Is there a newer version?

 

[JohnW2]

Yes I have had access to BSOL via my local library web portal. I had another search and have found the login page - http://www3.lancashire.gov.uk/onrl/BSI/BSiLogon.asp Just need to find my library card number now.

OK, thanks. When I have some time, I'll have words with my 'local' libraries (who claim I can't do it on-line), then. I suppose it's possible (probable) that different libraries (or LAs) vary in the extent to which they offer on-line access.

Found it - but the only 61439-3 is from 2012. ... Is there a newer version?

That seems to 'depend'. The below is the 'current' version being offered for sale by BSI. It's called BS EN 61439-3:2012, but, as you see, says it was published in March 2019 -so I guess probably includes contains Amendments. What's the 'publication date' on the version you're looking at?

Kind Regards, John

 

[mattylad]

Same as the one your looking at. (I didn't notice this before)

 

[SimonH2]

Many thanks, Simon. Very briefly .... Indeed, but the rest of that sentence reads
[QUOTE}[a consumer unit] incorporating components and protective devices specified by the manufacturer complying with
BS EN 61439-3, including the 16kA conditional short-circuit test described in Annex ZB of the standard.

... which seems to be a bit more specific.[/QUOTE]
Sorry I didn't pick up on this at the time.
You are correct, the option of a Consumer Unit with 16kA devices is one option - but that specific clause explicitly allows the use of an 61439-3 board with devices rated for the prospective faults current applicable to the installation. So in effect, there is the choice of "use a CU (with 16kA devices) and don't worry about fault breaking capacity" or pick any board as long as you've considered the ratings.
Effectively, it caters for the "painting by numbers" people who don't want to have to think about regs or calculations - just fit "standard" installations working to the rules of thumb and numbers in the OSG. I.e. the "installation means a CU, two RFC, two lighting radials, possibly one for smokes, and if the developer isn't a tight-a**e a radial for electric cooking" brigade.

The below is the 'current' version being offered for sale by BSI.

HOW MUCH Yes, without access via public libraries (or for some of us, a work subscription to the online service), it would indeed be a very expensive business getting these.

Checking online, the 2019 edition includes corrigenda from March 2016 and December 2015.

But then here's a thought ...
If BS 7671 says (to paraphrase) "use something to standard EN BS nnnnn", then does the electrician need to know what is in EN BS nnnnn in order to follow 7671 ? Or is it sufficient to simply take the work of the vendor who put "EN BS nnnnn" on the labels ?
I suppose the answer to that is generally no - you don't need to get and read the other standard unless you plan doing something "unusual" and need to understand the nuances.
Being a bit naughty, has anyone seen a CU described as "EN BS 61439-3 including 16kA tests to appendix ZA" ? I don't recall seeing such, but without such a statement, the person fitting it doesn't know for certain that the CU does in fact meet the requirements for that option - just having 16kA printed on the front doesn't meet the requirement since there are other tests in 61439-3 that it could refer to

 

[ericmark]

When I wanted a new distribution board (CU) for my house I simply went to local supplier and asked for a consumer unit for the number of fuses and MCB's I already had, only stipulation was all RCBO. He produced a unit with a SPD and asked if I wanted the SPD, I could see no good reason not to fit it, I looked at the RCBO boxes which were clearly marked type B, and mentally noted only really needed type A so actually better than required, and since I am not a scheme member I had the board fitted for me, cheaper than paying the council.

It was not until latter I realised all these type B RCBO's were in fact type AC curve B. I have two B32 type A on order, and they have been on order for some time now, intend just to change the socket in house RCBO's the rest I will leave as type AC.

But we do tend to trust the supply chain to supply the right gear, I knew I didn't need neutral switching as TN supply, and it seems the whole sale outlet, myself and the guy who fitted it for me all missed that the RCBO's were in fact type AC which is understandable since clearly marked on packet type B.

And that is what most do, we don't check the EN BS number we simply don't expect to be offered units that don't comply.

 

[JohnW2]

... Sorry I didn't pick up on this at the time. You are correct, the option of a Consumer Unit with 16kA devices is one option - but that specific clause explicitly allows the use of an 61439-3 board with devices rated for the prospective faults current applicable to the installation. ..

Indeed, but ...

So in effect, there is the choice of "use a CU (with 16kA devices) and don't worry about fault breaking capacity" or pick any board as long as you've considered the ratings.

I am obviously at a considerable disadvantage by not having seen the actual Standard but, my 'understanding'(based on other things I've read) is that the "16kA conditional short-circuit test described in Annex ZB" is a test that confirms that breaking capacity of a combination of 'standard' 6kA devices and the cutout fuse (hence, perhaps, the "100A" 'requirement"?) is satisfactory. Is that not correct?

I suppose a fairly crucial question is whether or not Annex ZB is 'Normative'. Is it?

But then here's a thought ... If BS 7671 says (to paraphrase) "use something to standard EN BS nnnnn", then does the electrician need to know what is in EN BS nnnnn in order to follow 7671 ? Or is it sufficient to simply take the work of the vendor who put "EN BS nnnnn" on the labels ? I suppose the answer to that is generally no - you don't need to get and read the other standard unless you plan doing something "unusual" and need to understand the nuances.

I think you've answered your own question, at least in terms of 'spec/characteristics of the product' (i.e. if it were a "Product Standard", which most Standards are) .....

In relation to such Product Standards (which is what we are talking about here), one surely cannot do (or be expected to do) anything other than to 'trust', and rely upon, the labelling (be it a CE mark, a BS number or whatever). Even if one has read the other Standard, it will usually be impossible (and not expected) for one to oneself undertake the tests to confirm that the product.

If, pretty unusually, it is not a Product Standard but, like BS7671 a "what to do" Standard, then things would presumably be different. If, in talking about CUs, BS7671 required compliance with BS WXYZ (another 'what to do' Standard), and if that Standard said that CUs had to be installed in fireproof cabinets, then presumably one would be expected to become aware of the requirements of BS WXYZ and to comply with them?

Being a bit naughty, has anyone seen a CU described as "EN BS 61439-3 including 16kA tests to appendix ZA" ? I don't recall seeing such, but without such a statement, the person fitting it doesn't know for certain that the CU does in fact meet the requirements for that option - just having 16kA printed on the front doesn't meet the requirement since there are other tests in 61439-3 that it could refer to

I suppose that depends upon the actual requirements of 61439-3 and, to some extent, as above, whether Annex ZB is 'Normative'. If the situation is that 61439-3 requires some 16kA test to have been passed (whether the 'conditional' test in Annex ZB or some other specified one - maybe in other Annexes?), then I suppose that compliance with the Standard is all one needs to know, and that the extra words you suggest would be unnecessary?

Kind Regards, John

 

[SimonH2]

Indeed, but ...
I am obviously at a considerable disadvantage by not having seen the actual Standard but, my 'understanding'(based on other things I've read) is that the "16kA conditional short-circuit test described in Annex ZB" is a test that confirms that breaking capacity of a combination of 'standard' 6kA devices and the cutout fuse (hence, perhaps, the "100A" 'requirement"?) is satisfactory. Is that not correct?

I suppose a fairly crucial question is whether or not Annex ZB is 'Normative'. Is it?

No, it's informative.
TBH it's "hard reading" But there seems to be two distinct test flows.
Where the device is rated for 16kA fault current, it's a simple test of applying such a fault and testing that things are still within spec after it's cleared - not set the cheescloth (draped around the outside of the enclosure to determine if excessive blast has escaped) on fire, clearance and creep distances not impaired, terminals not deformed, and still functional.
Where the devices are not rated for 16kA fault current, then there's further tests to be done, including on an outgoing device of minimum current rating.
The test rig does include a BS88-3 100A fuse, or a BS1361 type II fuse.

The purpose is stated as being :
The following test procedure is intended to verify the performance of the incoming device and its connections, and any other item in the CDB not separately rated in excess of 16 kA, when the complete CDB is protected by a fuse-link complying with BS 88.3 (formerly BS 1361). This type test shall be deemed to cover the use of any other short-circuit protective device having a Joule integral (I2t) and cut-off current not exceeding values given in item b) below, at the rated voltage, prospective current and power factor.

So as the old saying goes, every day is a learning day. I hadn't really picked up on the "combination including 100A fuse" bit.

But I am curious, does anyone ever find a prospective fault current ever approaching 6kA, let alone exceeding it in the real world ? I figure you'd almost need a large cross-section connection to the substation in your basement to get it. You'd need a Ze below 0.04Ω to get 6kA, if the distribution cable was (say) 35mm² in the street then that's only 30m needed assuming a zero impedance supply at the substation. You still only need 61m of 70mm².
Our service cable (i.e. just from our cut out to the street cable) at home probably provides half the required impedance on it's own.

 

[JohnW2]

Where the devices are not rated for 16kA fault current, then there's further tests to be done, including on an outgoing device of minimum current rating. The test rig does include a BS88-3 100A fuse, or a BS1361 type II fuse. The purpose is stated as being : The following test procedure is intended to verify the performance of the incoming device and its connections, and any other item in the CDB not separately rated in excess of 16 kA, when the complete CDB is protected by a fuse-link complying with BS 88.3 (formerly BS 1361). This type test shall be deemed to cover the use of any other short-circuit protective device having a Joule integral (I2t) and cut-off current not exceeding values given in item b) below, at the rated voltage, prospective current and power factor.
So as the old saying goes, every day is a learning day. I hadn't really picked up on the "combination including 100A fuse" bit.

I think that is the crucial bit. In my attempts to understand this, I've read some pretty complicated theory and maths about 'breaking capacity' when there is a combination of a fuse and MCB in series.

But I am curious, does anyone ever find a prospective fault current ever approaching 6kA, let alone exceeding it in the real world ? I figure you'd almost need a large cross-section connection to the substation in your basement to get it. You'd need a Ze below 0.04Ω to get 6kA, if the distribution cable was (say) 35mm² in the street then that's only 30m needed assuming a zero impedance supply at the substation. You still only need 61m of 70mm². Our service cable (i.e. just from our cut out to the street cable) at home probably provides half the required impedance on it's own.

Indeed. As you will have seen, I've often asked this question myself. Although my experience is pretty limited, I've personally never seen a PFC appreciably above 1kA, let alone 6kA - but whenever I ask the question, I am assured that some extremely low Zes do exist, at least in London. As you imply, I would have thought that, in most cases, the cable from the street to the cutout would often be enough, on its own, to get the PFC down to 6 kA or less.

Kind Regards, John

 

[davelx]

Measured Ze of 0.1 Ohms here in SW London.
TN-S system about 100m from 11kV substation.

 

[Adam_151]

I've only skim read this but few points in no particular order.

* Above 6Ka at a consumer unit would be possible, but unlikely, don't think I've ever seen it in a domestic environment, but have seen it at consumer units outside of a domestic installation
* Between 1 and 2 is usual, between 2 and 3 still possible, but less likely, above 3 quite unusual
* Slightly Erroneous to suggest that the 'devices' have a conditional rating of 16kA as per annex Za... its the assembly as a whole, it does then mean that if you use a motley collection of mixed devices you no longer have a 'type tested assembly' and therefore don't have a conditional 16kA fault rating....
* The familiar Wylex standard units would surely fail the test and leave a nasty skidmark on the cheesecloth if the plastic cover marked "FUSES" was missing, as they often are in the field
*The upstream fuse plays a significant part, just because there is a measured PFC of 16kA, the fault current on the downstream side will likely never reach this value, due to inherent inductance in the system, faults don't appear instantaneously, there is a very small rise time before it can be reached, this will be slowed even more as the fuse element heats up and resistance increases, and by the time it clears, its likely the full value hasn't been reached (look for 'cut off current' in datasheets). There was one manufacturer of switch gear who's 5A breaker was only rated at 1ka, but could be increased to 2kA with a fault current limiter accessory fitted to one of the breaker terminals. This was nothing more than a current path that had several very tight 180 bends where it doubled back and forth on itself in order to add a tiny bit of inductance so the breaker could safely trip on a fault that 'could' be upto 2Ka before it got there.

 

[mattylad]

BS7671 (as do others) refers to other standards in several places, IMO to fully comply you need to have these standards also (when they apply).
It would be a boring and probably not so useful exercise for someone to see how many standards are referred to and what it would cost for all of them, so you could say you have a complete set.

 

[JohnW2]

Measured Ze of 0.1 Ohms here in SW London. TN-S system about 100m from 11kV substation.

I continue to be fairly amazed. Even if the TN-S sheath had only the same impedance as the L conductor (which I rather doubt), that would presumably require at least 50mm² cable all the way from the substation to your installation. I must say that I had always assumed that the 'extraordinarly high' PFCs we hear about would/could only ever be seen with TN-C-S.

Anyway, even 0.1Ω only equates to 2.3kA, and we are talking about 'greater than 6kA' - which I find hard to believe would ever be seen in a domestic installation.

Kind Regards, John