I'm still working on the solar panels. They're not finished yet, but I'm making progress. This post is a bit about the design of the system.
As soon as you add a second panel to the system, there's a decision to be made about whether to connect them up in series (water goes through each panel in turn) or in parallel (there's a junction to split the water so it goes through both/all at once, then another junction to bring it back together again). Guidelines that I consulted on how to connect up solar panels tended to describe a combination layout, e.g. four panels arranged in two parallel runs, without discussing the pros and cons of each option. I had to do some general reading, and some thinking for myself, and here's what I came up with.
Panels in series make a longer loop for the water to go round, which means a more powerful pump is required. Panels in parallel require balancing to make sure one of the panels doesn't end up dominating the system, which the water essentially bypassing the others. I think that plumbers tend to favour parallel systems for these reasons, as balancing systems is something they do all the time, so it's no big deal to them. As for the effectiveness of catching heat from the sun, given a certain amount of sunlight falling on a given surface area, it might seem that there's no way that varying the plumbing could affect how much heat is captured, but there are a few factors that are worth considering.
Specific to my system is the fact that radiators have a high volume of water relative to commercial flat panel solar collectors. This means that they aren't so responsive; they take longer to heat up, so if there's only a short period of sunshine, they may not get hot enough to trigger the pump and take the heat into the thermal store.
Hot enough is an important point here: If my store is at 50°C already, the incoming water needs to be pretty hot before it's any use. This means I want to maximize temperature, if possible.
On the other hand, there's the question of panel efficiency. For all the insulation I've put in, they will radiate heat back into the air, especially through the glass. If heat can be taken away from the panel when it's only a little above air temperature, then very little heat will be lost to the air, so the panel will be operating at a higher efficiency than it would if it was a getting a lot hotter before the heat was taken away and stored. Now we have arguments in favour of both high and low temperatures.
Let's look at how serial and parallel systems vary in terms of temperature. In a parallel system, water in all of the panels heats up at the same time, starting cool then, if it gets hot enough, all being pumped into the store at once. In a serial system, water goes through each panel in turn, so the first panel might be considered a
preheat panel, and the last a
topping up panel. This means there are different temperatures at different points in the system, which could be a good thing, considering the conflicting requirements mentioned above. The first panel can operate at a lower temperature, and so be more efficient, and the second panel brings the water up to a temperature high enough to be useful, i.e. hotter than the water already in the store.
However, there is a slight weakness in the description I just gave you.
Water goes through means that the water is already flowing round the system. This won't happen until the last panel gets hotter than the stored water which, radiators being slow to respond, could take quite some time first thing in the morning. This brings us to the third panel.
How about having a
booster panel that will heat up quickly and get the water flowing round the system? There are many designs for DIY solar panels to be found on the internet, and most of them involve much smaller volumes of water than the radiator design. If I made a small one of these and added it after the bigger panels in the series, it would heat up more quickly when the sun hits the panels in the morning and so start the water circulating before the other panels had got up to temperature. After that, it should serve to boost the temperature as the preheated water flows through it, so giving the higher temperatures needed to be worth adding to the store, at the same time as allowing the other panels to operate at cooler, more efficient temperatures. That's the theory, anyway.
Having bent my pipe, I then used a 10mm steel bar - I have no idea where that came from - and hit it repeatedly into the aluminium sheet to make grooves. If you're thinking of trying this, it really is worth making a jig to shape the grooves. I started by setting the workbench to the appropriate width, but it kept opening out, which absorbed much of the energy I was putting in with the hammer.
The grooves in this aluminium sheet are carefully aligned to the bent pipe (visible in the background), not accidentally wonky
Once I had the two parts made, they needed connecting together. Following advice from builditsolar, I used silicone gloop to fill the gap between aluminium sheet and copper pipe. I then drilled some holes and tied the pipes into place with cable ties.
Each cable tie is looped twice over the pipe and joined at the back. I had hoped to pull the aluminium more tightly around the pipes, but the cable ties weren't strong enough.
All this got me as far as selecting a radiator for the other panels. I then selected a suitable sized window and made a box in the same way as I had for the other panels (I didn't tell you about the second panel. It was very much like the first one, except that the best fitting windows were wood framed house windows, which took a lot more work to prepare, being old and somewhat rotten). At some point before putting the insulation in, I realised that it would be a good idea to solder the connections that needed it before the pipes were snuggled up next to plastic insulating stuff. I'm glad I realised at that point.
Here is my finished booster panel, out on the hillside. The diagonal lines are reflections of overhead cables.
As you see, the window wasn't a very good fit for the collector, but it does allow space for a bleed valve to one side. The vertical pipe disappears under the top layer of insulation, to exit the box at the bottom.
And here it is in context:
So, I now have all three panels in position. Once I've connected them together I'll be able to fill them and start catching sunlight. That isn't quite the end of the project, though. Insulation is quite important, too, and it would be nice if the power supply didn't run across the hallway to reach the pump.