Switching sides - changing array generation panels (E/W orientation)

[the-chauffeur]

Sorry for the cryptic title . . . couldn't think of a better way to summarise what I'm asking.

Setup: even number of panels split 50:50 E/W (house almost exactly E/W oriented). Hybrid inverter and batteries.

Long story short, we've got 4 more panels on the roof than our inverter can handle; they're not connected up right now. To move past that elephant, we chose to put them there in the knowledge we were over-specing(?) for a few reasons; aesthetics to balance out the roof (esp. street-facing side) and longevity/expansion given that we couldn't guarantee they'd be available in 5/10/etc year's time. Using a higher power domestic inverter isn't really an option at this point (for reasons) but knowing that inverter/battery technology is changing all the time, we're also considering the longer term strategy of connecting them permanently when tech allows.

So today's challenge is whether/how we could incorporate the currently dormant panels into the current system without overloading it using switching. For the majority of its uptime, the sun is either on the east panels or the west, with both sets generating but the non-facing side generating substantially less. What would be ideal is if we could turn off two of the panels from the non-facing side whilst turning on the two dormant panels on the facing side depending on where the sun is. Not so much of an issue in summer, but very much more desirable as a system 'feature' in winter.

For the sake of this example, the array is split 10:10 with 8:8 permanently powered. In the morning, we'd want a split of 10:6, moving to an afternoon split of 6:10.

I've seen lots of stuff online about folks doing clever things with inverters using Home Assistant (we're in the process of setting that up) but much less on the manipulation(?) of panels. I suspect that what we're trying to do isn't that hard if we have access to the panel cabling and timer switches. Obviously the timer switching would need to be set up in such a way that there's no possibility of everything being on at the same time but beyond that, we're not sure whether there are any other limitations/pitfalls we might be missing.

Anyone done this sort of thing before and/or fancy chipping in advice, guidance, warnings and so on?

Hope that all makes sense - and thanks in advance.

 

[binky]

why can't you fit a bigger inverter?

 

[the-chauffeur]

As I said, reasons.

Right now the inverter is already at what I think is the upper limit of the domestic single-phase sizing (6kw). Future plans could include another, rather than a larger, inverter but that would involve some other control mechanisms to stop them fighting with each other. At the same time, generation volumes during summer are already, um, considerably higher than we're ever going to use.

Should probably add that our priorities are a bit different to most folks. We' don't want/need outright maximum return on everything all the time - at least not immediately. Like most folks, we don't use all we generate in summer and as yet we've not figured out how to sell back the unused stuff (and given what we can already put out I suspect there's a G99 certification in our future). This is a long game thing.

For the time being all we're looking to do is determine whether we can do something relatively simple to optimise the panel usage during the dark times. If we can't, the four dormant panels stay that way for the foreseeable and that's fine too.

 

[bladerunnerpv]

Which model of inverter do you have?
What's the spec for the panels?

 

[the-chauffeur]

As weird as this will sound, I've deliberately avoided putting in that detail because as I said, the question is about the practicalities of optimising a slightly peculiar array arrangement through switching. Without wanting to appear rude or difficult, I don't want to get into the size/scale of the kit itself.

If it helps, the array voltages are just within the upper limits of the inverter capacity/capability - I know 'cos I've asked the manufacturer. Right now, there's no more headroom to safely add anything without going outside the unit limits. To mitigate anything nasty when it's operating at the top end power generation-wise it's actively cooled so it sits comfortably at 10-15 deg cooler than passive equivalents.

 

[bladerunnerpv]

Presume your inverter has 2 x MPPTs, East side into one and West into the other?

You're trying to stay within the overall dc input watts from PV?

You'll have a max volts dc input spec for each mppt though, if you're at the max inverter limit with 8:8 then sounds like you would be exceeding the mppt input V limit by increasing the string from 8 to 10 panels? Not to mention voltage increase as temperature drops. Any opportunity to switch in/out panels would be limited by that.

Sorry I don't see how it could work in that case

Not to mention how the cables from the panels physically feed into the inverter.

If a bigger inverter is not an option then a second small pv inverter for the spare panels could be worth a look but the startup voltage could be a problem with only two panels

You might find the inverter has a high enough input pv dc current limit to re-wire the panels in parallel, 2x5 on each side of the roof, that would bring the string voltage down but increase the current. I think that's unlikely to work though, but check the numbers. You'll probably exceed the input pv power limit that way too.

Just my 2d worth, I have no practical experience on this...

 

[the-chauffeur]

Yup - 2 MPPTs - one E; one W. And you're absolutely right - the string increase from 8 to 10 would bust the limit given that it's a straight 8:8 split as it stands. As you say, re-arranging the arrays to a different set of splits might make it possible (and that's a big might) but equally, it's going to be so complicated and the benefit is likely to be so minimal that the maths and the cost of the wiring, switches and so on mean it's not worth the aggravation.

We've thought about adding a smaller unit and splitting the existing arrays but right now that's not really worth doing either, at least from a cost/benefit standpoint. When the second set of panels were installed (we added to an existing array), our installer told us the inverter manufacturer was working on a unit that would synchronise multiple hybrids, so if that ever becomes a reality, that may ultimately end up being where we go. Again, we're very lucky in that we're in no hurry whatsoever.

Thanks so much - that's the big (and obvious now you've pointed it out) flaw in what on the face of it seemed to be just a simple switch situation. It's an excellent point and if I'm honest, it's not the first time I've made that mistake. It's something I keep missing, which is why I posted here . . . and fortunately you've stopped me before I try something really stupid!

Thanks again.

 

[bladerunnerpv]

Cool, glad I've been useful for once!

Suggest you take a look at easypv, it's free and very handy for trying different options out. Pvgis is another option.

There are a few posts on the forum about two hybrids connected by an ethernet cable, it is possible in some cases, depends on the manufacturer and the inverter on a case by case basis It seems.

I think your simplest solution to use all the panels and feed into baytery would be to replace the inverter with a bigger one and submit appropriate Gxx form.

For the spare panels a small solis or victron pv string inverter might work, iirc they have low startup voltages, provided your electrics will support it. victron have a widget on their website which you put the panel spec in and it recommends an inverter

Like you say, it depends if it pays for itself though.

 

[binky]

if you fitted 2 change over switches and cable it right, you could turn on the extra panels in poor light / winter without too much bother. So basically run 8 panels for most of the year, 10 panels in winter, beyond manual switching it would involve relays and some controls, perhaps getting keen with a Raspberry pi to programme or Arduino, for which there some very keen forums. But as bladerunner says, you may run into problems with the MPPTs, panels tend to sit at max volts VoC before the system starts up and voltages drop as the MPPT adjusts to reduce voltage and up ampage to get power.

 

[arwooldridge]

I have two identical systems except that on in south facing and one is split EW (2mppts). I can compare the two systems directly and on sunny days the south facing one peaks at around 50% more power in the middle of the day, and the EW one just beats the S first thing and last thing, but the power then is very low anyway.
So yes I would say you can increase the number of panels on an EW system compared to what you would have on a S system. Provided it's total inverter power you are worried about (although in any case inverters just top out when presented with more potential power than they can handle, and most allow for 50% more PV power than the inverter can handle)
However you would need to check maximum voltages on the strings. You could get full power and voltage on one string at some point during morning or afternoon in summer.

 

[the-chauffeur]

Thanks folks.

It's clear the MPPT thing - and the specific/related capability of the inverter - is something I'd have to do a bunch more research on before I would be in any way confident that I wouldn't melt/blow/damage something. Making the switch during the hours of darkness would be simple/safe enough, but the midday switch is a whole different ballgame - and as you've warned, could end very badly depending on the inverter capability.

In order to get the right panels working at the right times, I think I'd need at least four switches so various circuits could be energised or disengaged. As a bare minimum, I'd need to split out/rearrange the existing string splits so they comprised half the panels on one side of the roof and half on the other. (as opposed to their current set-up of being split by roof halves). From there, it would be a case of creating sub-groups to energise/disengage when necessary.

I'm guessing Home Assistant could be used to control something that in turn would trigger relays to handle the switching. I've already put a load of time into sorting out the active cooling for the inverter and heating for the batteries and will almost certainly end up running all of that via HA. Either way, that's another rabbit hole I'm currently staring down and I can already feel that I've got some coding learning in my future.

Ultimately the maths plus the research plus the cost of rewiring will more than likely mean that for now at least this remains little more than a thought experiment . . . an interesting one but a thought experiment nonetheless.

Cheers!

 

[binky]

Simplist option is the 4 panels on their own inverter. It may not talk directly to the battery, but if they are feeding energy into the house, then your hybrid inverter should direct more energy into the battery as the house won't be demanding as much from the hybrid.

 

[Badger]

Or use 4* micro inverters on the ac feed direct to the mcb rcbo breaker in the consumer unit each solar panelnwould then run itself and youd only need a enlighten unit to get the info on the app

 

[Badger]

 

[Badger]