Crofter

I'd like to add a simple wall mounted panel heater to our holiday cottage, so that we can keep it at a reasonable temperature when nobody is there, using the off peak overnight rate. At the moment we leave the heated towel rail on when the weather is cold, which is probably not the most cost effective method. I was thinking of just buying a hardwired wifi relay (e.g. a 25A one from Shelly) but then I came across heaters with Wifi functionality built in. That's quite appealing because it would allow people in the cottage to use the heater as well. I'm a bit worried about the longevity of any built in 'smart' technology though. Having owned an early Smart TV which after a couple of years was no longer supported, I feel that it's easy for the app to be updated and suddenly older tech no longer works. Maybe this could happen to the Shelly relay too, but at least in that case I only have to replace that one small component. Any thoughts? And any recommendations?

This seems like a good time to point out that the Scandinavian countries seem to get on just fine heat pumps too.

Sadly you are correct. A house has to be in budget, in the right location. For a lot of us anything more than that is a bonus. It's not like buying a car where you can shop around all over the country.

I've never been a huge fan of subsidies. In fact other than getting cavity wall insulation in my previous house, I've never claimed a grant or subsidy for anything, ever. Which is pretty unusual when you own a croft! Make electricity cheaper, put those costs on to fossil fuels. That's all they have to do. Oh, and get rid of standing charges. Spread those costs so that those using the most electricity pay a bigger share.

And the biggest ticket item is the triple lock. But pensioners actually get out and vote, so there's no way any realistic government is going to touch them.

On the ceiling, I used 100mm PIR. I marked the rafter line in places that wouldn't be covered by the PIR, then ran a straight edge down to draw that line on to the PIR. It helped that I was using paper faced, not foil. I can't remember quite how I managed to drive the screws in square, possibly I used a block of wood or something to act as a set square. You only need to be super accurate in one plane, you can be pointing a bit too far up or down the line and still hit the rafter. The insulation washers are essential. I think I was using Pz3 screws, at these lengths that what you'll find. I was fortunate and got most of mine on clearance from Screwfix. If you do mess up and miss the rafter, just try again. And if you end up leaving a hole, you can foam it up. My battens went directly over the same line. Really, the first set of screws just need to hold the weight of the PIR for a short while until the battens go on. I used 25mm battens which is minimal really in terms of creating a service void. It worked, but it made the wiring pretty fiddly. The back boxes I was using were designed for a deeper cavity and I had to trim the lugs on them to make them fit. That got a bit tedious after a while!

I used very long screws to hold the PIR to the rafters directly. No battens attached to the rafters. I used insulation washers, which are big plastic things, to spread the load and stop the heads pulling through. I then installed a polythene vapour barrier, with battens following the line of the rafters, and even longer screws. The PB then went on top of those battens.

I know diverters aren't the right answer for everybody. My interest in this was actually sparked by seeing a cheap used diverter for sale locally. The unusual thing about my use case is that, being a holiday house, I'm not in a position to load shift. Guests are going to use electricity when they want to, not when the sun is shining.

I looked up the Eddi and the ability to power two different devices sequentially looks absolutely perfect. Is this a common thing on diverters? What should I be considering when choosing the heater? As it happens I never removed the old storage heaters when I put in my A2A, so it would be tempting just to wire up the one that's in the hallway, which would add a nice bit of background warmth to the whole house. But I don't know what wattage is appropriate for this, e.g. should it be sized to take account of the array output?

My heat pump is air-to-air, the water is simply heated via immersion in a UVC. So the array is just going to boost that when there's enough sun. There are two immersion elements on the tank, I'm not sure how best to set everything up to take best advantage of the diverter. I guess that ideally I would have the immersion powered by the diverter set to a higher temperature, and the one powered by the grid would be set lower as a backup. But of course one of them is half way up the tank. My thinking with the panel heater would be simply to dump excess power, but I'd really want it to operate only from the solar. A very crude way of doing that would be to simply fit a timer, so that it is always off overnight. But it would be far better to set it up to only come on when the diverter had already fully boosted the UVC. I'm not sure how to do that.

That's a very similar application to mine, great to hear that it's working out so well for you. Using a panel heater as a dump load is an excellent idea- and I presume that its own thermostat is enough to prevent overheating the property? My understanding (and I could be wrong) is that a grid tied inverter feeds in a higher voltage than the grid, so that the house draws preferentially from it- and any surplus just gets exported, whether you are getting paid for that or not. So e.g. I could leave a panel heater switched on and it would take from either the grid or the panels depending on how sunny it is. Obviously not a very clever way to do things, but I'm just trying to understand how the system will work.

That's basically what I was suggesting, and how I built my own house. Except I used 2x1 battens on top of the PIR. Make sure you mark the line of the rafters before you cover them up.

This might not be possible, but another thought: In the same way that the PV diverter heats the water automatically, could a system be put together which would switch on the heat pump if certain conditions were met? (Sufficient solar input, tank topped off, temperature of house below X°). It's a Panasonic A2A unit, installed last year. I'd imagine that such a task would, if even possible, require a high degree of geekiness involving things like Pi or Arduino etc, not that I really know anything about those. Maybe it would be better to consider adding a battery instead?

If the goal is to add extra insulation, could you use rigid insulation boards that go across the rafters? This would also reduce thermal bridging.

I thought I'd need a second inverter because the extra three panels will see quite different conditions. Maybe I'm overthinking it. Tbh they don't add a huge amount, it would be easier just to stick to the main array of six.

OK I am settling on a design. Assuming that I can erect 12m2 of ground mounted PV within permitted development, I'd be putting up a structure that's 3m tall at the highest point, and which would have six 590w panels along the top at a 39° angle, plus another three horizontally along the front at 74°. I've arrived at these figures after spending a few hours farting about with PVGIS and I think they're optimal. It does take me slightly over the 12m2 footprint but I don't think anybody would notice, and if they did, I'd just have to bring those three extra panels all the way to vertical. I'd have a 3.68kw inverter for the top panels, and a 2kw one for the extra three. The array would be mounted out behind the house, at a distance of around 20m. Where's the best place to put the inverters? I could see it being fairly easy to build an enclosure under the panels. Then run SWA down to the house? Are there any rules that need to be followed about building the structure? I could obviously just use wood but perhaps unistrut would be longer lasting. Inevitably I'd be very tempted to box it in underneath, lots of people seem to do that. If fully boxed in, it would look something like this:

Most panels seem to operate around 32-38v, and the grid-tie inverter I was looking at has a minimum of 60v, so yes you'd need two. But as I understand it you could wire up pairs in parallel, so long as the current didn't exceed the specs on the inverter. By the way I'm starting to mull over the possibility of switching tack and installing a system on the other house instead. I could just go ground mounted for optimum performance. It's a slightly bigger tank and the bills are generally higher, so it's probably more worthwhile. I read somewhere that permitted development for ground mounted solar in Scotland was to be increased from 9m2 to 12m2, but I haven't seen anything definitive about whether that change has come in to force.

I installed A2A last year and I'm very happy with it. The main unit sits under the floor with 200mm insulated ducting running to the larger rooms, and 150mm to bedrooms. The outlets are at floor level (plinth in the kitchen, just above skirting elsewhere). It's inaudible unless you put your ear right up to it. The MVHR I fitted makes more noise.

As I've said, I'm familiar with off grid systems, but I think the principles are the same. Panels in series will run at the current of the lowest output panel on that string, and at the voltage of all the panels added together. Panels in parallel will run at the voltage of the lowest performing panel and the current of all panels added together. Panels in series give a higher system voltage, allowing smaller cable sizes and generally better efficiency. But shade on any one panel will drastically cut output of the others. Panels in parallel don't get the benefits of higher voltage, but they function independently and shade on one panel doesn't drag down the others. Panels put together in a string will act like one big panel, subject to the above rules. There's nothing to stop you from making up all sorts of combinations, so long as you stay within the specifications of the receiving charge controller.

OK I get it now, I looked at the spec of that Solis and it's limited to an input current of 14A. The panels I'm looking at have Imp of 13.5, which seems awfully close. So any attempt to wire in parallel would go way above the max current.

What prevents you from combining the inputs in parallel? Sorry if I'm missing something obvious. E.g. on my old boat I had two 260w panels in parallel, I just ran them to the same input terminals on the MPPT. I had another MPPT which had four 100w panels all in parallel. On a boat parallel is the way to go because of variable shading, the panels can end up with quite different levels of output.

So I looked up that Solis and it's got a 60v startup. I'm used to boat stuff which will kick in at much lower voltages. So 60v means I need to put two panels in series to get above startup. Presumably that means I could have a pair of panels on each elevation, but I'd ideally want separate inverters for them? Nothing is limiting me to 3kw, I'm just trying to keep this system small, simple, and low cost. Anything above what's needed to heat the UVC is likely just going to go to waste. As a holiday house, it is unoccupied throughout much of the day, and it has low heating demand. And as a non MCS install I wouldn't be getting anything for export. I'm a bit sceptical that the current export rates are sustainable tbh and I don't want to build a system based on them.

Good point. I'm playing with PVGIS and wondering about putting two panels on each aspect of the roof. I hadn't considered that this could complicate things. I've heard of micro inverters, would those be useful in this situation? Or, can I just run panels in parallel? On my boat I have always done that, but that's a smaller system.

I doubt I'd bother with secondhand panels, they're the cheap bit after all.

Well that makes me feel a lot better actually. I had my own idea of what to do and every passing joiner who popped in for a nosey said I was doing it wrong...