Multiple Transfer Switches in same enclosure (Domestic Battery Install)

[wildej]

Hello all,

Just a quick caveat to say that while I'm posting this information here in DIY, it will in fact be installed by a professional sparky. However, the slightly irregular nature of the installation means that I want to field a wider pool of ideas to find both an acceptable but also optimal (in terms of tidiness) solution before I get it done for me.

This relates to a domestic battery installation (using a Victron Multiplus) and specifically dealing with three components:
1. A DP Isolation by means of MCB/RCD or RCBO to feed the AC from the grid to the Inverter/Charger;
2. A generator transfer switch to allow a CU that includes critical loads to be fed either from the ACout on the inverter or from the grid;
3. A second gen transfer switch to allow a separate CU that includes non critical AC loads to also be fed from either of ACout or the grid.

To avoid address any concerns, the second gen transfer switch will rarely if ever be on anything other than grid. The only reason for doing this is because the existing PV is grid tied and comes from that second CU so in the event of grid failure, the PV will cease to function (unless it come on to the inverters little island). I also don't want to simply put everything through the inverter because it is limited to 50A even on pass through and I don't want to constantly doing active load management to prevent overloading the inverter, which would be the case.

This arrangement requires three devices to be fed from two sources. The RCBO and both of the transfer switches need a connection to incoming grid. This could be easily achieved with a Henley Block of course. Likewise, the two transfer switches need to be connected to the inverter ACout, which again could be achieved with a Henley Block. However, the net result of this is three DIN rail components, all separately enclosed and couple of Henley Blocks with associated tails all connecting together. Doesn't sound very clean or tidy.

In an ideal world I would drop all three of the components in to one enclosure, leaving things much tidier. Two things concern me:

* Is there any particular regulatory reason why these separate units should not share an enclosure?
* How can this appropriately and safely be achieved in a single enclosure?

I've attached a basic diagram to help anyone kind enough to assist. The inverter/charger can pass-through and output 50A, so I've been I been robustly assuming 16mm cables. In this case, this would be mean that each of the connections has to be 16mm as well (possible scope I guess to reduce these of course). In the diagram I have illustratively shown the connections being achieved by doubling up the connections to the devices, but I'm not really anticipating it being practical, desirable and sensible to try and cram two 16mm cables in to single screw terminal. So how best to do it?

I'm aware that in a conventional domestic CU, there will be an earth and neutral screw terminal block and the line would be on a bus bar. The layout of the devices I think precludes the use of the bus bar and most CU's I've seen don't tend to have screw terminal block that could take multiple 16mm connections? Even if they did, how good/bad an idea would it be to reuse the earth and neutral terminal blocks as neutral and line terminal blocks for my use case? I'm less familiar with modern metal CUs, but I presume these are earthed so potentially directly connected to at the earth terminal blocks, which would not be ideal for me.

Ultimately my problem is about splitting one in to three and one in two, which could be achieved by using 4-way and 3-way terminal block connectors respectively, but even to my unqualified mind that seem a sub-optimal solution to have floating about inside an enclosure. Can you get DIN rail terminal block connectors?

EDIT [More googling less typing and I would have seen that DIN Rail Terminal blocks would very much do what I need - so much a case of should I do that]

I would really appreciated and throughts/suggestions on both a clean, tidy and most importantly safe and compliant solution.

Thanks in advance.





John

 

[binky]

A grid tied inverter will not function wothout the grid, unless you can manually set it to 'island mode'.

Most inverters do have a UPS funtion, ie a seperate AC outlet that can take a limited load for critical applainces like a freezer.

Daft question, but what are you trying to achieve? Do you suffer regular power cuts, or just like the idea that shuld one occur then you can be off gid?

 

[wildej]

A grid tied inverter will not function wothout the grid, unless you can manually set it to 'island mode'.

Most inverters do have a UPS funtion, ie a seperate AC outlet that can take a limited load for critical applainces like a freezer.

Daft question, but what are you trying to achieve? Do you suffer regular power cuts, or just like the idea that shuld one occur then you can be off gid?

Sorry - should have explained. The inverter/charger in question here is a Victron Multiplus-II which is able to operate as a battery charge/inverter connected to the grid, by means of a bi-direction AC connection (ACin). It also has the scope to operate in back-up mode whereby critical loads connected separately (to a terminal referred to as ACout1) can continue to function from the inverter in the event of grid failure.

The Victron will 'pass-through' up to 50A from ACin to ACout, so critical loads get grid feed when available. In the event of grid failure, the Victron has an automatic relay disconnect of ACin (the island mode your refer to, to be G99 compliant) and also relay based neutral-earth bonding for RCD functionality (separate earth required of course).

My house has two separate CU due to a garage/extension. These are currently connected independently via Henley Block to meter/fuse. It just so happens that one of these CUs carries the majority of the critical loads but which are cumulatively less than the 50A limit on the Victron for passthrough and/or output. These will be by default my critical loads.

I need to be able to isolate the Victron and it requires an RCD being directly connected to the mains (hence the DP RCBO). The first transfer switch is also essential, as I need to be able to switch the Critical load circuit back from being on ACout (pass-through) to being directly connected to mains - to cover the eventuality of maintenance or failure of the inverter/charger. So two of these components are necessary anyway.

However, the existing grid tied solar PV inverter (which as you say requires 230v 50hz to function) sits on the second CU (not the critical load one) and therefore is on the grid side of the VICTRON (essentially only ever connected via ACin) and will be not operate in the event of grid failure. The aim of the second transfer switch is to be provide a functionality to allow the second CU to switch over to the ACout side of the Victron and thus be added to the loads that can be run within island mode. Most importantly, this would meant that the Solar PV inverter would benefit from functioning on the island allowing the batteries to be charged from the solar. To clarify, I can't 'move' the solar PV because it needs to stay where it is to be FIT compliant.

In answer to you question about grid failure - yes I do appear to have a grid that fails quite a lot and frankly, call me nuts, but I think this is going to be even more common in the future (that is the pessimistic in me speaking). However, I don't want to adopt the approach of putting all my house loads through the ACout because without constant active load management by everyone at home, it would be all to easy to exceed to the 50A passthrough. It is quite practical to do this for short periods of time though.

That was a very long explanation for why. My main question was how. Having looked around at DIN Rail Terminal Blocks like the Wago Topjob, these seem to be way to go - inside a suitable enclosure with suitable long rail and space for the modules. However - is there a reason I shouldn't be asking a sparky to do this. I can imagine them coming up with lots of reasons why not, including 'i've never done that before', but are their real reasons?

 

[binky]

The Victron will 'pass-through' up to 50A from ACin to ACout, so critical loads get grid feed when available. In the event of grid failure, the Victron has an automatic relay disconnect of ACin (the island mode your refer to, to be G99 compliant) and also relay based neutral-earth bonding for RCD functionality (separate earth required of course).

so it has a UPS function, same as a lot of other inverters. Would it not be easier just to wire critical loads to this output, or is that difficult to achieve / insufficient for your needs?

One thing I will say, if you are pulling highish loads, you will need an awfully large battery.

Din rail enclosures, perfectly acceptable, I like to avoid jamming components in to give them space to 'breath' and avoid overheating, but if rated for the ampage there's no real reason for doing that. You will find most domestic sparkies are not familiar with this type of equipment as it's the sort of work that you get in industrial places/ large commercial buildings. Those wagos look rather nice, a lot better thn what I used to work with.

 

[wildej]

so it has a UPS function, same as a lot of other inverters. Would it not be easier just to wire critical loads to this output, or is that difficult to achieve / insufficient for your needs?

One thing I will say, if you are pulling highish loads, you will need an awfully large battery.

Din rail enclosures, perfectly acceptable, I like to avoid jamming components in to give them space to 'breath' and avoid overheating, but if rated for the ampage there's no real reason for doing that. You will find most domestic sparkies are not familiar with this type of equipment as it's the sort of work that you get in industrial places/ large commercial buildings. Those wagos look rather nice, a lot better thn what I used to work with.

Battery is scale-able. Starting with around 10kwh but can increase modules to suit. Most critical loads are separated, but I want to give myself 'switching' capability, rather than rewiring. I'm assuming (although this is always debatable) that this is notifiable, so I would prefer to get it done once and certificated.

Your last point is my issue. I realize that using din mount terminal blocks and/or distribution blocks aren't completely conventional in a domestic environment and the natural reaction might be to stick with what they know (although I haven't asked yet, so shouldn't imply that this will be the reaction). I just prefer to try and get close to what I think I want first and then discuss with the sparky.

What I would prefer is a single (if necessary large) enclosure with DIN rail to encompass the three main components referred to above as well any of the terminal/distribution blocks for connecting. This should allow for a tidy finish and this single enclosure can then be clearly labelled as the point of isolation and transfer associated with the battery storage.

The wago's look nice, but I am concerned with the overall depth as they are front cable entry. I don't think I could position these on the same single DIN rail of say a traditional CU enclosure and have 10-16mm cables managed on the front entry? May have to use the traditional top/bottom entry screw type.

All good fun. Thanks for you help.

 

[binky]

I would buy an enclosure from wago if they do one, just to ensure they fit properly with the lid.

New circuits are notifable in domestic works.

 

[wildej]

I would buy an enclosure from wago if they do one, just to ensure they fit properly with the lid.

New circuits are notifable in domestic works.

Couldn't seem to find anything that would cover off a situation where I want access to switches on the front of the panel, but also have enough depth between the front of the enclosure and the rail to allow front entry cables that obviously don't come out of the front of the enclosure or require the cables to bent aggressively.

I was instead thinking of doing the attached. This puts the RCBO and two transfer switches in a generic circa 12 module CU enclosure (good quality of course) and uses two Techna DB35 DIN rail mounted distribution blocks (which are top/bottom entry like an normal MCB/RCD etc) for splitting the LINE connections. It then uses the two NEUTRAL screw terminal blocks that are generally in a CU anyway for splitting the neutrals (but keeping them separate as is necessary for the separate DP isolation). Obviously need to earth the metal enclosure as well.

What do you think of the attached? This seems to me to work in practice. It provides a single 'distribution panel' for all of the isolators associated with the battery storage in one place, that can be easily labelled and understood. It removes a whole load of single tails that would need double isolation, clipped to the wall linking separately enclosed devices and switches.

However, does it fall foul of some regulation that I'm not aware of (quite probable giving my lack of knowledge)? I'm aware that it is good practice to generally only use compatible devices (MCB, RCDs etc) in a CU, whereas this mixes different devices. However, this seems to be done commonly anyway and this isn't actually a consumer unit installation in the conventional sense, just a enclosure for distribution switches.

Thanks again!

 

[binky]

looks fine to me. Mixing devices is the subject of much discussion on another thread, and shouldn't be done

 

[ruggedscot]

Ok.... not too sure about the RCD - This is feeding the victron ? the victron feeds two selectors to select source either victron or mains...

Where is the neutral earth link ? how is that maintained ? lost that and then you are working how the RCD protection downstream will operate.... needs further thought

 

[wildej]

That is covered internally to the Victron. When the grid is lost, an internal relay disconnects the inverter from the grid feed in order to be G99 compliant and operate in island mode (to ensure no chance of backfeeding and making DNO/DSO cables live when they shouldn't be). When this happens, a few ms later there is a second internal relay that bonds the neutral to earth, allowing the RCDs downstream to continue to have an earth referenced to neutral. The earth being an extra electrode to replace the TN that has to be assumed to be lost with the LINE and Neutral in event of grid failure.

I have since realised that one of my CUs is a split RCD type and has some circuits unprotected (except for MCBs). As a result, I was planning to add a type S time delay 100ma RCD between the Victron ACout and the transfer switches, also in this enclosure, as effectively in island mode it will be TT earthed and everything must be RCD protected. In due course, CU will be changed and have RCBOs, but not yet.

Thanks for your reply. Hopefully that makes sense? What did you think about the single enclose point though, doing all these tasks?