Plug in generator installation connection point

[JamieP]

Proposed is to mount a changeover switch on a wall in a switchroom. The generator inlet would connect to this. The change over switches supply from a normal supply submain to the generator and then feeds a subboard.

4c+e normal supply, 4c+e from generator inlet and 4c+e to DB

Can this be done compliantly? I was under the impression all generators had to connect to a MSB but can't find the info to back it up. About to refresh myself on 3010:2005 but just wanted to seek some better guidance.

[AlecK]

There is no rule prohibiting connection of an alternative supply at a DB, to supply all / part of that DB's load.
It is well-established practice for large installations with a centralised alternative supply to distribute that supply via (essential power" submains to those DBs that supply essential loads, and to have contactors in those DBs to transfer essential loads only to the alt. supply.

Note that for non-permanent generators; the genset N must be switched, as the inlet N-pin is otherwise an accessible live part.

While 2005 edition of "3010" is the one cited in ESRs, that citation (ESR 60](2)] is only applicable for mains-parallel work.
ESRs impose no particular requirements for alternative supplies.

Accordingly for general guidance you should be designing / installing in accordance with Clauses 7.3.2 & 7.8.4.1 of "3000" (cited edition).
These clauses require compliance with "3010;" but all such citations within "3000" are to latest published edition of the relevant Standard, not to (in this case) 2005 edition of 3010.
So you should be looking at 3010:2017 for guidance that is considerably more up to date, and clearer, than 2005 edition.
As example, that requirement to switch the inlet N is explicit, whereas in 2005 you have to deduce it by first recognising that it's an accessible live part.
Also in 2005 whether to use 4-pole switching for the normal supply is left open, while 2017 discourages switching Ns unless there is good reason to do so.

2017 edition includes some NZ-only provisions for connection of genset at a swbd without a N-E link (ie most DBs).
Essentially setting up the genset supply to replicate the normal (TN-C-S) supply;
by installing a dedicated set of N&E bars and an earth electrode that are only in service when the genset is the selected source.
These are well illustrated in the relevant Figs; but - as always - you need to read the clause [2.7.10.2] in order to understand what's shown.

If that doesn't appeal, you could adopt a "Part 1 solution" and follow the Australian clauses / Figs instead; as being both no less safe and in accordance with long-established NZ practice.

Regardless; you need to consider how to provide fault protection.
We normally provide this by automatic disconnection of supply, using an overcurrent device.
The same device we also use to provide overload protection and short circuit protection.
This system relies on a high earth fault current to cause operation of the device.
Low-powered sources (regardless of genset, inverter, or tx) cannot supply sufficient fault current.
The logical alternative is to use one or more RCDs instead of overcurrent devices for this purpose.
Especially if any of these essential circuits are required to be disconnected within max 0.4 sec.
With a 3-phase genset, you may have sufficient fault current available to use over-current devices as normal; but should check.
Especially as, being a plug-&-play set-up, you have NO control over the capacity of the actual genset that will be used, other than it is probably intended to be able to supply the connected "essential" load. Whether the genset that actually arrives has that capability will be subject to availability at the time.

Same consideration means you need to consider that whatever genset is connected may not have overload protection suitable for your inlet conductors. Accordingly Section 4 of 2017 edition requires these conductors to have their own overload protection, thus not relying on whatever may be included in the genset.

Plus you need to label the inlet to avoid use of a genset that has the winding connected to frame, as most gensets do.

[JamieP]

Hey Alec,

I will have a proper read but just to not this situation the genset isn't connected directly to the DB the proposal is to have an intermediate change over switch mounted on the wall between both switchboard but not directly run to either incase that changes anything. So it's just an independent changeover on the wall.

[AlecK]

It's still switching supply to load supplied from that DB between 2 sources; one being "normal supply" and the other being an "alternative supply".
Physical location of the source selection device makes no difference to how the requirements apply, but will affect the detail of how you comply with them in a particular case.
Where many people go wrong is taking a short-cut by following a Fig without bothering to read the requirements the Figs illustrate. Then sometimes can't find a Fig that exactly replicates their particular situation.
In this case, you'll be applying section 2, and also Section 4 because it's a plug-in genset - remembering that this may modify requirements of Section 2.
The Section 2 Figs will illustrate the various typical arrangements that apply for any genset.
The Section 4 Figs will be consistent with these; but are there mainly to illustrate the special requirements for detachable connections.
All are "typical", so the source selection is shown within a switchboard.
So there won't be any single Fig that shows all of what you need.

Make sure you have 2017 edition including A1, as the original 2017 edition included many errors in both text & Figs.

[JamieP]

I see 2.5 (g) but I guess that is referring to when the generator is supplying the whole installation? In this case it's supplying a part installation do not relevant?

The main part I'm struggling to wrap my head around is your comment around needing to bring in it's own MEN/electrode.

If a 3 pole/4 pole change over was used wouldn't the generator just still use the MEN and electrode of the upstream MSB? Fault path for earth would just go back via the earth in the cable between the change over and the MSB, through MEN link then via the same mentioned cables N through to the generator N since that feed N isn't getting switched?

[AlecK]

Yes 2.5(g) sets a default of only supplying installation (or part) from genset via MSB;
but immediately allows an Exception provided 2 conditions are met.
As published (using "and"), both conditions must be met - making it completely pointless.
Was supposed to be "or" - but no-one (including me) spotted this error in time to get it fixed.

By now we should have had a further amendment, but first Covid and then re-structuring of EL 001 got in the way.
The old subcommittee looking after "3010" is no ore, and new "Maintenance Team" for it has just been appointed.
Most of the members of the old subcommittee are no longer in place; and there's no proper record of the issues that were pending - that list needs to be re-compiled and await next full revision (no more Amendments unless for urgent safety issues).
Meanwhile almost nobody on NZ has looked at it, because it's not the edition cited in ESRs and most don't understand that the secondary citation directs us to latest edition. So almost everyone just follows the older edition, and all is well enough as no safety issues arise and no complaints of non-compliance are filed... and if they were, chances are vanishingly slim that EWRB would understand about secondary citations either.

Now it's a race between getting necessary changed made and MBIE maybe, eventually, getting their finger out to update Schedules 2 & 4

Back to 2.5(g):
Condition (i) is about co-generation. Basically anti-islanding; just as we all expect with inverters.
But by using "and" this condition makes it impossible to ever use a genset for alternative supply to a DB; because as soon as normal supply is lost the genset has to cease supply also.

Condition (ii) is what matters; and is the option for alternative supply.
Step 1 is disconnecting the normal supply.
Step 2 is having a MEN earthing arrangement.
Step 3 is controlling EPR.

Which brings us to your 2nd question.
For Step 2; yes we could continue to use the installation MEN (at MSB); as you suggest, and as the Australians do.
Except that 2.7.10.2 prohibits doing so.
While it's under the heading of "changeover device in the generator neutral"; it goes much further; and gives the detail of how Steps 2 & 3 of item (ii) if Exception to 2.5 ("earthing arrangements") must be carried out.

The fact that we have this detail, along with the related Figs, shows absolutely and beyond all argument that the "and" back in Exception to 2.5(g) is an error; as it was not intended to absolutely prohibit connecting a genset as alternative supply at a DB. If that had been the intent, then option (b) of 2.7.10.2 (ie connecting at a swbd that doesn't have an MEN link) would not exist at all.
The point is not to prohibit such connections, only to restrict how this can be done. So we can proceed on basis that this is an "or" choice between following (i) for supplementary supply, or (ii) for alternative supply.

Where we want to connect to a swbd with no MEN link; we are forced to follow Figs 2-15 / 2-16.
These Figs show the used of 4P / 4P source selection, and the setting up of a separate N&E bar for the genset supply on the genset side of the source selection device.
The reason for 4P / 4P switching is to avoid this N-E link ever being on parallel with the installation's MEN; because that would lead to load current flowing in the submain PEC.
The background is 'explained' in the Notes to the clause

Which is in itself a case of unacceptably poor drafting, as a Fig that is labelled "Typical" cannot impose any requirements. Requirements are supposed to be set in the text of clause (according to SA's Guide for drafting Standards).
Similarly the Notes can't set requirements. So a good case can be made that none of this is actually mandatory.
In which case all we need to comply with is the three steps of the Exception to 2.5(g).

However just as we don't expect EWRB to understand secondary citations, we can't expect them to ignore the clear intent that this complex arrangement be installed.

Ostensibly to provide both the MEN- type earthing configuration (converting the IT supply from genset to TNS for the installation);
which it does, but is not needed for.

And also ostensibly "controlling" earth potential rise.
Which it actually can't do, because if anything happens to cause EPR while the genset is supplying the essential load, it will either
a) come into the installation via mains N, across the MEN, and onto the installation earthing system - to which the essential loads are still connected; or
b) come up through the two earth electrodes, and again raise the potential of the entire earthing system.

The one situation this additional N-E link does anything useful for is if an earth fault occired somewhere in the load and the installation's MEN were open circuit while that essential load was being supplied from genset. As, for example, during the replacement of MSB, or while testing is being carried out on the normal supply.
In absence of the extra N-E connection; the fault current path woulf be from point of fault, vis subcircuit PEC to DB, via submain PEC to MZB, across MEN, then back to source (genset) via submain N and genset N.
So without the MEN, fault protection for the essential load would not be provided.
The low probability of this actually happening doesn't mean it can be ignored; but whether this complex arrangement is the best way of dealing with it is questionable. Perhaps the risk could be managed by requiring a label at MEN, warning to not open-circuit it without first disabling the genset. Similar to labelling used where supplementary supply is fed-in at a DB under 4777.1; to warn those working on 'upstream' swbds that there may be back-feed

The electrode part of the arrangement serves no useful purpose at all.

Note 4 states that the (optional) 100 mA RCD provides earth fault protection and also EPR protection.
True for fault protection provided either N-E link is in place; but only where the genset can't provide sufficient fault current to operate overcurrent devices, and can do this without the extra bars & electrode .
Not true for EPR; other than WRT the (extremely short duration) EPR that results from high fault current flowing in a PEC. Which is why we have max operating times for fault protection . The RCD does nothing more for that source of EPR, and nothing at all for any other source.

You may well ask how we ever ended up in this situation.
The answer is that if NZ practitioners can't be bothered to read draft Standards and submit Comments on them;
there's a risk of the requirements being dictated by a tiny group. In this case 1 of only 2 Kiwis on that subcommittee.
Australians of course generally don't care enough about NZ-only provisions to get involved. Polite of them, but not very helpful.
So one man was able to ram through his concept of how things should be done.


Where to from here?
Up to you, as you're the one who will have to certify the work.

Option 1: follow & certify to 2005 edition.
- technically not compliant with ESRs, but who will challenge it?

Option 2: follow the NZ provisions of 2017 edition.
- extra trouble & expense, and (because of that "and") strictly compliance with letter of Standard is not possible; but compliance with intent is.

Option 3: (assuming non-domestic) adopt a Part 1 solution, and follow the Australian provisions of 2017 edition.
- requires issuing of a Certified Design; but not hard to show no less safe than following Part 2 down the special NZ provisions.

[JamieP]

If setting up the temporary MEN and electrode when on generator supply. Do you believe this constitutes a main earthing system and in turn becomes high risk PEW?

[AlecK]

Good question.
The arrangement meets the definition of "main earthing system" in ESRs; so yes installing it is high-risk PEW and so requires inspection.