Battery backup earthing - PME supply

AJJewsbury: 
 

Sorry Andy, whilst this would at first appear correct, you also need to disconnect the supplier's neutral from the installation for a number of reasons. Failure to do so can render RCD protection inoperative (or could cause nuisance tripping) … and that may well be important because of the reduced EFLI in island mode.

I disagree that the entire installation needs to be disconnected from the supplier's N - rather just the portion that's supplied by the embedded generation needs to be galvanically isolated. Just as if we had a UPS that was serving just a few critical items - the remainder of the installation remains connected to the (dead) grid supply during power cuts.

   - Andy.

Yes, that is true, the Neutral must be disconnected from any part of the installation supplied by the backup.

Not all UPS work like that, as we discussed in another thread recently. If you take your supply at LV, you can't rely on the distributor's N-E link, and of course you can't just “earth the neutral” upstream of the UPS to fix that without disconnecting the Neutral.

So in that respect, UPS vs backup are identical.

The IET CoP covers both options - “backup” for part of an installation, and “backup” for the whole installation. There still has to be a means of switching L & N to the distributor's supply from the backup system, and forming N-E connection.


Anyway, the point I was making, that others make to me, is that you're not “isolated from the distributor's Neutral” if you keep the distributor's means of earthing connected - it's the Neutral conductor that needs separating, NOT the whole system from the distributor's means of earthing (and neutral).

Sorry Andy, whilst this would at first appear correct, you also need to disconnect the supplier's neutral from the installation for a number of reasons. Failure to do so can render RCD protection inoperative (or could cause nuisance tripping) … and that may well be important because of the reduced EFLI in island mode.

I disagree that the entire installation needs to be disconnected from the supplier's N - rather just the portion that's supplied by the embedded generation needs to be galvanically isolated. Just as if we had a UPS that was serving just a few critical items - the remainder of the installation remains connected to the (dead) grid supply during power cuts.

   - Andy.

This is quite a good link as it indicates the actual output is limited to half the battery capacity as a product of discharge current and voltage (ie 0.5 C if I'm using the correct terminology) and it would appear that the warranty is either 10 years or earlier if the declared kWh charge/discharge cycles have been used up…..

I don't know the UK price but it would be good to work out the cost per kWh extracted from the battery using the declared capacity and 90% DOD cycles compared to present domestic costs of around 18p per kWh.

Regards

BOD

Some manufacturers do offer a complete solution that meets all the requirements in the UK.

Sometimes, this can be easier than trying to effect an installation from a kit of parts.

As we move into a situation where we have a number of “SMART” or “green energy” products and so on, all of these disparate items need to work together as a complete system.

The setup and integration of the management systems may well be more difficult than the “electrical problems” (I use quotes here because, of course, they are only “problems” for domestic-scale installations because they are new concepts … those familiar with large infrastructure, healthcare, etc. installations will already be familiar with supply change-over and battery-backed supplies).

gkenyon: 
 

There is no problem having the DNO's earthing terminal connected to MET at the same time as your own earth electrode (just like extraneous-conductive-parts). In fact, a consumer's supplementary earth electrode is recognized by BS 7671. See Fig 3.9 and the descriptions below it.

What you MUST NOT do is connect N to PE within your installation (downstream of the DNO's service head), without disconnecting the distributor's neutral first - regardless of whether the supply is TN-C-S (PME), TN-S, or TT. This is not only an issue for protective devices, but also potentially contravenes ESQCR.

Ah I see, so you’re saying the order of change over needs to ensure line conductors are broken before the N-E link is made by the inverter, and then broken first before the DNO supply is reinstated?

I would hope (wrongly or rightly) the inverter has been designed to take care of this, where it’s “backup” output shuts down, breaks the N-E link and then reinstates what is DNO supplied/synchronised power onto its backup output. 

I can see some inverters on the market have separate automatic change over devices using sets of contactors, however they don’t seem to address the issue of the N-E bond. 

This is something from Solax - install guide at the bottom of the page.

https://www.wattuneed.com/en/mounting-and-accessories/21133-x1-eps-box-solax-0768563817090.html

My thinking here is that actually this device breaks the grid input before it allows connection of the inverters backup supply. I think a simple relay could be added here to make a E-N link on switch over. 

There is no problem having the DNO's earthing terminal connected to MET at the same time as your own earth electrode (just like extraneous-conductive-parts). In fact, a consumer's supplementary earth electrode is recognized by BS 7671. See Fig 3.9 and the descriptions below it.

What you MUST NOT do is connect N to PE within your installation (downstream of the DNO's service head), without disconnecting the distributor's neutral first - regardless of whether the supply is TN-C-S (PME), TN-S, or TT. This is not only an issue for protective devices, but also potentially contravenes ESQCR.

It's absolutely fine for the appropriate switching devices (island mode isolator to disconnect the distributor's live conductors, all lines and Neutral and the N-E bond relay to form TN-S when the grid is fully disconnected) to be part of the inverter, battery management system, etc., provided they meet these requirements.

You will, of course, need to provide the additional consumer earth electrode if one is not already present. The value of electrode resistance required will depend on the size of RCD used for the island-mode circuits ("essential loads") and whether there are any other complications such as TT-d outbuildings or EV charging points.

You must ensure that protection against electric shock and overcurrent are available in all modes of operation.

Because an EESS charges the battery as well as as discharging it, you will need to check the rating of the CU is not exceeded. For example, if the CU is rated for 100 A, and there's a 100 A service fuse, and a 16 A output battery storage system - by feeding 16 A in at one end through an OCPD, because that OCPD gets hot it contributes to the total heat load in the CU - therefore the CU should be rated for 116 A.

All of this is covered in the IET Code of Practice for Electrical Energy Storage Systems.

This area seems to be a minefield of potential complications.

Z.

gkenyon: 
 

AJJewsbury: 
 

This would then form an island TN-S system for the inverters output?

Correct. Typically all the "Earths" are (DNO and local electrode) are permanently connected together - the DNO's can't be relied upon during a grid failure and the local rod becomes just another bonded extraneous-conductive-part when the grid supply is working. But in island mode all the live conductors (N as well as L) of the inverter are kept completely separate from the DNO's.

   - Andy.

Sorry Andy, whilst this would at first appear correct, you also need to disconnect the supplier's neutral from the installation for a number of reasons. Failure to do so can render RCD protection inoperative (or could cause nuisance tripping) … and that may well be important because of the reduced EFLI in island mode.

Further, the DNO won't like it if they isolate the PME main for maintenance, and potentially find voltage on it they weren't expecting.

Regardless, it's effectively a requirement of ESQCR.

Hi,

If the inverter makes a direct connection between earth and its “backup“ neutral output, the disconnection of live and neutral “backup” outputs would already be isolated to prevent feedback into the inverters “grid“ input (otherwise the grid input of the inverter would become live powering non-essential loads). 

So the inverter itself (at least the type I have posted) will disconnect incoming L+N from its battery back up output. 

If we installed a change over switch that disconnects L+N into the inverter, or indeed the whole installation, there would still be a permanent connection through the earth conductor (as this has been switched in automatically) on the intertevers backup output and the N-E link on the PME system (for example on the cut-out). 

Albeit the L+N can be isolated firmly, there will also be a ”link” of sorts between the PME connection, the ground rod and the N connection of the inverters output.

The inverter is designed to provide an instantaneous change over to prevent loss of power to essential loads, so the use of a manual change over is not required. 

What impact (if any) will the PME earth (inside the cut-out or the street, wherever the DNO may have done this) to neutral link have, with regards to isolation? 

Thanks. 

Indeed, as Andy suggests, earths solidly strapped, and all pole change-over (Ls and N ) from DNO to inversion, so load N is lifted off the DNO N during island.

M.

AJJewsbury: 
 

This would then form an island TN-S system for the inverters output?

Correct. Typically all the "Earths" are (DNO and local electrode) are permanently connected together - the DNO's can't be relied upon during a grid failure and the local rod becomes just another bonded extraneous-conductive-part when the grid supply is working. But in island mode all the live conductors (N as well as L) of the inverter are kept completely separate from the DNO's.

   - Andy.

Sorry Andy, whilst this would at first appear correct, you also need to disconnect the supplier's neutral from the installation for a number of reasons. Failure to do so can render RCD protection inoperative (or could cause nuisance tripping) … and that may well be important because of the reduced EFLI in island mode.

Further, the DNO won't like it if they isolate the PME main for maintenance, and potentially find voltage on it they weren't expecting.

Regardless, it's effectively a requirement of ESQCR.