OK, I'm working on the premise:

1) Panels are cheap and it's best if you install all you can at the start while there's scaffolding etc.

2) Ideally I would like to be off-grid, but PV production in the middle of winter is b....ks, so I'll still use a grid connection then. But I would like to maximise PV production in the shoulder seasons.

3) I have a 1P / 100A grid connection to UKPN.

So, I'm planning about 48kWp of panels split across 4 roofs (in 2 pairs) in various orientations. And then:

1) Asking the DNO to allow a 12kW 1P inverter with grid connection for the house (ASHP, DHW with immersions, kitchen etc) which will support 24kWp of panels on the house (2x clipping). This will be integrated single stack (probably Sig) with as yet unknown amount of battery.

2) Having a separate 12kW 1P for the other 24kWp of panels on the 'cart lodge' but disconnected from the rest of the house circuits and the grid. This will only be used for car charging.

What do you think?

Edited yesterday at 14:491 day by Alan Ambrose

Firstly, with a 100A connection why start with the premise of only exporting 12kw? At least ask the utility if they can support higher export? Then it's simple and easy. Even if they limit export below generation it seems like keeping things simple would be better. Having said that, you might not even get 12kw so that then creates another problem for you.

I'm not sure your approach does anything more than make things complex. A 24kwh battery is not enough to fully charge a car and even if you are mostly just topping up you will likely come across times when you empty the batteries in the off-grid system before you charge the car. Also, if the off-grid system had spare solar stored in it's battery but the house did not, you would have no way to use that capacity. So I would look for everything to be grid connected (with ability to continue operation without the grid). That way you get the full capacity of all batteries available.

If you have significant excess generation that can't be exported what are you going to do with it? Seems like a huge waste to spend a fortune on install and not have anything to use it for.

If you plan a workshop with potential for 3p machines then an alternative setup would be to use inverters to generate the 2 other phases (all connected to the same battery). 1 inverter is connected to the grid, the other two phases purely rely on solar/battery (inc charging from the grid connected inverter). I doubt this is a common option but I know you can do it with some inverters (Victron). Still I'm not sure it makes much sense from a financial POV.

Just over clock the inverters, fit PV to run at max amps or max voltage. We have 6.5kW PV on a 3.6kW inverter and 2kW on a 1kW inverter. Any excess is just clipped, but rarely does it clip anything.

But if going with hybrid inverters and battery you can put as much PV as you want on the DC side as long as you stick each array in it's own MPPT.

Battery and inverter I would investigate if you are running on a cloud and walk away if you are. Our GivEnergy is a good example it can be run in full local mode, so when GivEnergy went belly up everything just carried on working. A cloud based system you stuffed. Sig I believe need a web connection to have full functioning - so if that's the case, I would look elsewhere.

Also if you want full house functionality in a power cut, look for a system that has a gateway, that will allow full inverter output to house with grid off. Most normal hybrid just give a small ups output which is great if you just want a light on.

@Alan Ambrose

Have you done a decent usage analysis yet?

You may find you need less than you think.

>>> Battery and inverter I would investigate if you are running on a cloud and walk away if you are.

Good point - looks like Sig needs a connection every 90 days.

>>> Have you done a decent usage analysis yet?

Ah, ROI is not the objective - the objective is to be as nearly off-grid as possible.

>>> If you plan a workshop with potential for 3p machines then an alternative setup would be to use inverters to generate the 2 other phases (all connected to the same battery). 1 inverter is connected to the grid, the other two phases purely rely on solar/battery (inc charging from the grid connected inverter).

Yeah I do have a 3P lathe :) I wrote off this option though as, possible technically, but too tricky for UKPN / my chosen sparks to get their head around.

>>> Firstly, with a 100A connection why start with the premise of only exporting 12kw?

Good point, I believe the max on 1P is 17kW. I think Sig's max 1P inverter is only 12kW though.

>>> A 24kwh battery is not enough to fully charge a car

Yeah, I was planning on only charging direct from PV - it seems overly complex and expensive to me (I know that's ironic) to charge a static battery and then use that to charge a car's battery. That means either charging from the sun during the day or from the grid when the sun's not shining. That means having the car charger connecting either to the off-grid inverter or the grid via a (probably manual) changeover switch.

1 minute ago, Alan Ambrose said:

>>> If you plan a workshop with potential for 3p machines then an alternative setup would be to use inverters to generate the 2 other phases (all connected to the same battery). 1 inverter is connected to the grid, the other two phases purely rely on solar/battery (inc charging from the grid connected inverter).

Yeah I do have a 3P lathe :) I wrote off this option though as, possible technically, but too tricky for UKPN / my chosen sparks to get their head around.

I'm not sure why UKPN would need to know. Only the first phase inverter would be grid connected. The other inverters would share a neutral but not be connected to the mains line voltage at all.

Agree finding an electrician who is comfortable might be challenging. But maybe look up what’s involved in doing it as you might want to keep the option open for future and that might impact the design

1 minute ago, Alan Ambrose said:

>>> Firstly, with a 100A connection why start with the premise of only exporting 12kw?

Good point, I believe the max on 1P is 17kW. I think Sig's max 1P inverter is only 12kW though.

Why limit yourself to one inverter on grid? Nothing to stop you having multiple. Just need to be setup the right way so they honour the export limits. (There are simpler and less simple ways of doing this depending on design).

1 minute ago, Alan Ambrose said:

>>> A 24kwh battery is not enough to fully charge a car

Yeah, I was planning on only charging direct from PV - it seems overly complex and expensive to me (I know that's ironic) to charge a static battery and then use that to charge a car's battery. That means either charging from the sun during the day or from the grid when the sun's not shining. That means having the car charger connecting either to the off-grid inverter or the grid via a (probably manual) changeover switch.

And what you propose isn't complex? I think it's worse. Solar only wouldn't work. You need the battery (or live connection to grid) to smooth solars variation. Many car chargers can be set to limit output to approximately match what is coming in via solar while still being connected to the grid, so just use one of them with a normal grid tied system.

I think you should step back, start from the basis of sizing the inverters to be big enough to power the house + be able to export the maximum allowed/generated. Any bigger than that is pointless. Ensure you have enough MPPT's to enable proper segmentation of the roof (limit shading impacts). If there is a mismatch look at alternate systems/arrangements (eg, Victron that decouples MPPT's from inverters - also sells combined models). Work out the amount of generation you expect from the solar at the darkest time of year. You have an awful lot planned so it's likely that you will still generate a meaningful amount. Direct attention to areas that may lower demand to match this winter generation as much as possible and then look up data to see what to expect in terms of multiple cloudy/sunless days in a row and determine how much battery you need to ride through that. Get a car charger (and immersion controller) that can adjust output based on solar generation.

Obviously your plan doesn't make much financial or environmental sense, but if you want to do it that's the approach I would look at.

I should note, that if you plan to own an EV and keep it relatively charged up then the need for big batteries diminishes if that car has a built in ability to export it's energy. Many cars these days can output 2-3kw that could be directed to a generator input on the appropriate inverter (if it exists) or a battery charger direct charge from the car into the house batteries (easy if 48V system, likely not possible on a Sigenergy system). Having said that, I think Sigenergy are one of the few companies that claims to be able to directly interface with the DC side of the EV battery so could take charge that way.

It is both cheaper and more environmentally conscious to use the EV battery as the long duration emergency backup as it saves oversizing the house battery for the extremely rare case. The car doesn't need to be able to support the max load of the house. Just trickle (3kw) charge the batteries over many hours. That should get you through most extended outages when combined with even limited solar output (worst case scenario you conserve energy a little to extend the time). Advantage of intentionally building such a setup is that you end up with a very large portable battery. So in the extremely extreme event you have the potential to get in your car and drive somewhere else to charge it and then drive back home where you can use that energy.