AJJewsbury said:

I was thinking of the wording of the NOTE to 530.3.201 - which seems to say to me that no marking of devices to the standards mentioned in Appendix 1 always implies bi-directional. e.g. Section 531.3.3 calls up "BS EN 61008 series" and appendix 1 lists "BS EN 61008" without any date qualification - which I think could be read as the NOTE applies to any version of BS EN 61008 - even pre 2017 ones.

The NOTE itself says 'standards as listed' so it could be argued that the assumption is only valid for the dated version as listed.

As far as I know, the standards referenced for RCDs, circuit-breakers, RCBOs and AFDDs that bear the same reference have always had the requirement since the , although not all versions of predecessor British Standards (with a different reference number) did. However, there may be an inadvertent reason marking is not present (see below).

AJJewsbury said:

I suppose you're going to point out that the more specific standard references in Appendix 1 are all date stamped though (e.g. BS EN 60898-1:2012+A12:2017) -  and so the NOTE could be read as applying on to that version ... so maybe there's some ambiguity there...

I would not even go there ... see above, perhaps also 511.1 and the fact that it can only be the intent that a version of a standard pertains to the referenced versions of standards it references, but would also point out the following:

• BS 7671 does not apply to products themselves, only their selection. A NOTE in BS 7671 cannot dictate what is and has been in product standards.
• A standard cannot be 'interpreted' by another standard. Only a court can interpret a standard. The NOTE is to be taken for what it is ... a note for you to check elsewhere, something that's recommended (but you are still responsible for conformity) etc.
• An error could have been made by a manufacturer, or for a particular batch of products, or the marking may have been inadvertently removed. Looking for the absence of something is never a good way to 100 % verify something. It would still be an assumption?

we have RCD testers, it would be relatively easy to hotwire an rcbo with flying leads from a suitable mcb so that it is not connected to the busbar, wire it both ways up and do ramp and time tests to see if it trips in each direction. 

"During an eicr I have come across a wylex wrds 40/2 BS EN 61008 (RCD) Residual Current Device on a board for the solar inverter."

Why was it installed, what purpose does it serve? Does the solar inverter circuit need RCD protection?

gkenyon said:

or the marking may have been inadvertently removed

Deliberately, even.

I must say that no markings = bi-directional seems most unsafe. Presumably, at one time, nobody even thought about it.

gkenyon said:

Looking for the absence of something is never a good way to 100 % verify something

And just to clarify, before someone asks, yes I also mean when we try to "prove dead". I used the term "prove dead" on purpose here, because we are not, in fact, doing that, but rather "checking for the absence of voltage" ... this does not (always) mean that the system is completely "dead", just that you can't detect voltage at that point in time.

'Absence of evidence is not evidence of absence.'

I must say that no markings = bi-directional seems most unsafe.

Presumably it's an attempt at consistency and backward compatibility - since the original passive RCD designs (with the test button connected between N on one side and L on the other) we naturally reversible - and as most designs were used both in the UK and Continental Europe - where the convention for CU layout is the reverse of ours - with usually the bus-bar (or lots of wire links) above the devices and the outgoing connections below, there would have been a natural market pressure to allow devices to be connected ether way around (not that that's precisely the same as having power on both sides when the device is open, but that's another debate...)

   - Andy..

OlympusMons said:

we have RCD testers, it would be relatively easy to hotwire an rcbo with flying leads from a suitable mcb so that it is not connected to the busbar, wire it both ways up and do ramp and time tests to see if it trips in each direction. 

That won't necessarily identify full bidirectional functionality. In some cases, the device is only damaged when the test button is pressed, for an incorrect orientation of connection. There may also be issues with the way arcs are extinguished, if the device is required to break high fault currents "the wrong way", and this orientation has not been type-tested according to the relevant standards.

gkenyon said:

Looking for the absence of something is never a good way to 100 % verify something

And just to clarify, before someone asks, yes I also mean when we try to "prove dead". I used the term "prove dead" on purpose here, because we are not, in fact, doing that, but rather "checking for the absence of voltage" ... this does not (always) mean that the system is completely "dead", just that you can't detect voltage at that point in time.

'Absence of evidence is not evidence of absence.'

What is your standard of proof? More likely than not, or beyond all reasonable doubt?

I think that I prefer the latter for proving dead.

Proving a negative has never been easy. If I claim to be able to run a mile in 4 minutes, and then do so (most unlikely) I have proved my ability, even though I might not be able to repeat it straightaway.

However, if I claim not to be able to run, how do I prove it? You might very reasonably say that I am just not putting in any effort.

Chris Pearson said:

What is your standard of proof? More likely than not, or beyond all reasonable doubt?

Legal proof vs scientific proof is a good question - legal has two (one for civil, one for criminal).

Scientific enquiry uses a slightly different approach. Philosophical truth is perhaps another thing.

Engineers have to balance all of them

Chris Pearson said:

I think that I prefer the latter for proving dead.

That's interesting ... and it depends on how you define "dead". It's entirely possible to 'prove dead' using a voltage indicator and traditionally taught methods, and have 'danger' present (as defined in EAWR) if there is a minor and seemingly inconsequential action that doesn't involve operating isolators, or switches. Removing an earthing, bonding or protective conductor connection under conditions of diverted TN-C or TN-C-S system neutral currents is a good example ... but there are other conditions that might lead to failure to detect the potential danger ... perhaps we could term these 'errors or omissions', or simply 'never thought of that'.