|
Post by aerojon on Jul 25, 2011 6:44:14 GMT -5
Hi there, I've been meaning to get something up here for ages (since getting so much good advice from Cye) but there are never enough hours in the day! I've made mistakes and learnt some stuff while putting up some panels and I hope some of this helps someone to avoid making the same mistakes in the future! I'll post bits as time allows, but here's my starting point. - Existing "boiler house" (a shed!) with a wood pellet boiler and 1000 litre accumulator tank (with solar coil)
- Pitched roof on shed with south-facing elevation (no accident!)
- 20 x 70mm indirect solar tubes (i.e. heat-pipe rather than fluid within the tubes), bought from ebay for £130. Seller disappointed...
My plan was to install the panels on the shed roof, directly above the cylinder. The advice from Cye was to use 10mm copper because of its lower fluid volume (compared to 15mm) and ability to bend by hand. However, because I've recently plumbed most of the house I had lots of 15mm fittings, tools, pipe etc, and the pipe runs are very short so I decided (and Cye agreed) to stick with 15mm. I bought one of Cye's "little brown" 12V pumps, and a simple control cicuit from REUK (http://www.reuk.co.uk/buy-SOLAR-PUMP-CONTROLLER-CIRCUIT.htm), along with thermometers, pressure relief valve and assorted other bits and strung them together with my fingers crossed. System schematic to follow... I hope... The executive summary after about 2 weeks of operation, half of which was the gloomiest, cloudiest summer week in the history of the world, is that it works! But I hadn't got the air bleeding right and cooked the wires leading to the panel temperature sensor... which stopped the pump... which made the situation worse! I don't know what temperature it finished up at (I was out over midday) but there was a lot of melted pipe insulation, and when I got back the panel thermometer (which seems to under-read a bit) was showing 96 degrees C. Ulp! Air-filled pipes don't absorb heat quite as well as water-filled ones! Got to pop out now but if anyone has any questions about my incompetent installation, please ask! Jon.
|
|
|
Post by aerojon on Jul 25, 2011 11:38:40 GMT -5
Here's a basic schematic for what I've done. I've misused some of the shapes (Visio didn't have a shape for "glass tube solar panel", would you believe?) but I hope it's clear enough. It's not to scale but is fairly accurate in the relative sizes and positions of the components.
Not shown is the filling loop (which just pipes in mains cold water to the bottom of the circuit to fill and slightly pressurise it).
Also not shown are two clip-on pipe thermometers, because I move them around a bit. One's somewhere on the exit from the tank and one's somewhere near the exit from the panels.
Problems with bleeding the air suggest that I might have failed to make the horizontal pipe runs slope slightly upward towards the bleed valve at the top. It's also occurred to me that the horizontal run by the pump would have trouble allowing air to rise towards the pump when it's running because the pump's pushing fluid back towards it, so that might need thinking about.
As I've been having such problems with superheated pipes I've left the bleed valve slightly open to allow any air to eventually bubble out. The amount of water lost is quite small and I'm keeping the system pressure topped up so whatever's lost is replaced.
Obviously this isn't an ideal arrangement, so I think I'm going to add a couple more bleed valves - one at the tank end of the horizonal run from the pump and another at the top of the return from the panels, as it exits the left panel. Although a properly built system shouldn't have need of automatic air vents, I'm going to make the external ones automatic so I don't have to keep climbing up on the roof!
I might have run the pump dry at some point because it's a bit noisier than before and I know the bottom bearing needs water to lubricate it.
I had a big problem with the pump controller unit, when the panels got very hot. The controller wasn't running the pump even though the panel thermometer was reading 96 degrees (and that was a couple of feet away from the panel itself). It was only when I removed the sensor from the pipe that I saw that the wire insulation had softened or maybe even melted, and bare wire was exposed to the pipe where I'd cable tied it on. The sensor was thus shorted out and the control unit got a misreading. In fact it might have done damage to the circuit as according to REUK it would have supplied 12V to the microcontroller instead of the 3V that it would normally supply. Looks like I'll have to keep that wire away from the pipe next time!
For now, I'm running the pump direct from the power supply which is itself on a mechanical socket timer. The timer only has 15-minute resolution but that seems to be good enough for now. I've got it on for 15 minutes every half hour from 9am when the heat arrives, working up to permanently on at 1pm and down to zero again at about 6pm. Not the most efficient controller in the world but I'm getting some useful heat and my panels aren't boiling!
It seems to me that there's a very fine line between working panels and Windscale-type meltdowns. Is this just because I'm bad at plumbing or does everyone have these issues?
|
|
|
Post by aerojon on Jul 25, 2011 11:39:36 GMT -5
Hang on a minute, where's my attachment? I'll try again. Attachments:
|
|
|
Post by cye on Jul 25, 2011 14:00:22 GMT -5
Hi Jon. Thanks for writing up your progress. Only a few teething problems to go.
£130, now that's a steal!
The sensors are a weak point of the REUK kit, agreed. Conor (aka caveman) suggested some types of heatshrink are high temp rated and you may wish to try insulating the wires (i.e., if the sensors still work!)
The REUK controller uses LM335Z temperature sensors which output a voltage which is proportional to its temperature - 2.73V at 0 degrees C, plus 0.01V per degree above zero. Therefore at 25 degrees, the sensor outputs 2.73 + (25 * 0.01) = 2.98V. Senors are <£1 apiece on aBay.
If something else has blown on the controller, Neill at REUK may well do a repair job for a few quid. I've only spoken to him on the phone once a few years back but he seemed keen and proud of his work, so worth a try!
Few points that may help you:
[1] Pump should go on the cold side of the loop, pushing water from the coil back to the panel. Fit your pump as low as possible in your solar loop.
[2] Never use automatic air vents unless these are isolated by a ball valve or gate normally turned off. This is because, when your panels stagnate, as they inevitably will, steam and/ore fluid will otherwise vent through the open AAV and you'll lose fluid from the circuit. As a closed loop circuit is supposed to automatically recover, you do not want to allow anything to escape other than air at initial setup.
[3] don't leave any air vents open, even manual ones. If you do this then you don't have a pressurised system, your pump will not work correctly and water may not rise to the highest point in your system. Instead, at all high points in your loop, fit manual air vents with some collection and storage capacity using some 'standing pipework', say 3-5 " vertical 15mm pipe teed off your main loop, ontop of which your manual airvent is sited. the standing pipe will collect air for you to be vented periodically. in a closed circuit you will have a lot of manual bleeding to do for the first day (or first few hours in hot weather as dissolved air comes out of solution more when hot). then close your airvents and check periodically. once you've the air out it shouldn't get back in.
if you really need to use an AAV on your shed roof, isolate it with a ballvalve with a lever on it. I'm sure you can devise some sort of remotely operated mechanism (cable + spring?, or pole?) to temporarily allow venting.
what pressures are you running your system (pipe loop) and expansion vessel at? if the relative pressures are not set correctly your pump will not work properly and your system will not recover properly from stagnation.
hope this helps
cye
|
|
|
Post by aerojon on Jul 26, 2011 9:33:55 GMT -5
Yes, the more water I boil the more I think these panels are really pretty powerful and I got them way below market price. Very low fluid volume as well, which seems to be good from all perspectives. I did have some heatshrink but not the high-temp stuff. I'll definitely have a look at that - good idea, caveman! I had to laugh as I read your "Solar Pipework - General Guidance" - I thought, "was this written about me?". Melted grey pipe insulation? Check! Solder melting point - blimey, I'm glad I only had lead-free to hand but even that might not be enough. By chance I'd got HT insulation for the external pipe runs, but only because it seemed more weatherproof. I wish I'd read that first. I'm still unsure about the "upward sloping pipe" question. My intuition is that everything should slope up towards a bleed valve if possible, but I read a discussion here www.aecb.net/forum/index.php?topic=2063.0 that says "..leading the pipes downwards from the manifold ... so that gas does not preferentially rise into the pipe." which seems to be the reverse. Are they saying that we should *aim* to have air and steam collecting in the manifold? I was also wondering whether I should be tilting the panels slightly up from one side to the other, so that air can travel through the manifolds rather than sitting in them. I've got them flat at the moment, and the AECB discussion seems to be saying that this isn't a problem (but I don't understand why). As for your other excellent advice: [1] That makes perfect sense now you've said it. I had originally planned to place it on the cold side but there was something about the arrangement of the inlet and outlet ports that made me think it'd be easier to site on the hot side. At that point I wasn't anticipating such extreme temperatures in the loop, and I definitely didn't give enough thought to air/steam traps. Will move the pump down this evening if I get time. [2] Yes, stupid of me. Again, I was thinking that steam would be a rarity, or rather something that I should design against. Turns out it's pretty much an inevitability and of course I don't want to be venting steam all the time; I want to keep that precious hot stuff in! [3] Good, I got something right - that's what my only air vent looks like at the moment - 5" vertical-ish pipe with a vent at the top. I've been doing lots of venting today to save the poor pump pushing steam and had a nice Ivor The Engine moment this lunchtime when I loosened a compression fitting at the panel exit and released a cloud of steam. Very impressive from a solar power point of view but not great safety or efficiency. Can I assume that most people vent air when the system's below boiling point? The system pressure varies at the moment with all my venting, but I would normally have it at about 1bar cold and it might rise to 2.5 bar when very hot. Does that sound reasonable? I don't get the "relative pressure" part of your question - what relative to what? Do you mean the "rest pressure" (or pre-charge pressure) of the expansion vessel versus the system pressure? If so, then I'm not sure! I'll have to check that - there's a schrader valve on the bottom so I can stick a pressure gauge on it when I drain the system to move the pump. Thanks for all your advice - you just might save us from an explosive / melting solar heating system. Jon.
|
|
|
Post by aerojon on Jul 26, 2011 12:07:17 GMT -5
Ok, so my expansion vessel is at 1.4 bar now the system's been depressurised. Does that sound reasonable?
|
|
|
Post by cye on Jul 26, 2011 14:44:16 GMT -5
if your EV (air side) is at 1.4 bar when there's no pressure on the other side, then pressurise your loop to 1.6
1bar EV air to 1.2bar system is also ok.
check out the narrow boat thread, where you'll see a hand-drawn schematic with some ideas for easy partial drain-down of the system to allow for easy reading (say annually) of the EV air airside pressure.
don't be critical of your own work, you're nearly there! bleeding air at initial setup is always a pain ('bleeding air'!). the 'pro' commercial guys use a £500 external pump affair with an air trap to help them out, us mere mortals need to design our solar loop plumbing a little more cleverly than the commercial guys to allow for DIY bleeding and maintenance.
i read the link you posted. your intuition is correct. not sure what the guy means really. just make sure there's an airvent up at the panel and at every other high point in the loop. i also put one right beside the pump too for good measure!
What's power output are your panels rated for?
|
|
|
Post by aerojon on Sept 18, 2011 14:15:41 GMT -5
Hello again Cye, I can't believe it's been nearly two months! I think I'm done with hospital for a while so it's time to get back to solar thermal investigation. First, to answer your last post: Thanks, I've got the EV at 1.0 bar and the system at 1.2 bar. Might as well keep the pressure as low as possible I think. You're very kind about my airlock-happy system. I've rebuilt it following your advice (schematic to follow) with the motor at the bottom, a bit of careful pipe-slope-engineering (i.e. bending the horizontal runs by hand to make sure that they slope upwards) and the addition of a second bleed valve adjacent to the panel output. There's still the occasional burble of air when the pump stops but I'm getting there. I have no idea what power output my panels are rated for! They were imported direct from China by an HVAC company and by the time I got my hands on them (ebay) there was no documentation at all. There's no maker's mark on any of the parts and no way to identify anything. It seems like a prototype really. All I know is that the tubes are 70mm, which is pretty large from what I've been reading in these forums. But now, the main reason for my post - graphs! I love to know exactly what's going on and there's no substitute for real-time temperature graphs. Here's the spoiler: tempmon.jordansplace.co.uk - temperature versus time for the boiler room ambient temperature, tank bottom, middle, top and average, and solar supply and return at the tank. I'll be adding another two sensors at the hot and cold ends of the solar panels when I'm fit enough to get on the roof. I started to detail how I'd set up the graphing but it got a bit long so I'll put that in a second post. The insights that I've gained since setting up this monitoring have been hugely useful and I'd recommend it to anyone. I found that the tank was losing 1-2 degrees per hour whereas it was supposed to lose that in a day. My intuition was that the wood boiler might be to blame, because it circulates water between itself and the tank even when not lit in order to measure the tank temperature and decide whether to fire up or not. The boiler is fairly well insulated but even so it's a bit of an Aga and gives off loads of heat to the boiler room. I turned the boiler off and consulted my graphs and to my delight the steep decline in tank temperature was halted! To add to my joy, the tank average temperature saw its first upward movement due to solar since I'd installed the monitoring equipment. I knew that solar gains need to be jealously guarded but I'm even more paranoid than ever now I know (or think I know) that my wood boiler was squandering my meagre gains. Cye had told me that I might need to slow down the pump to maximise efficiency. Well, the graphs seem to support this idea - the hot input from the panels can be seen to fall very rapidly when the pump comes on, then rises when the pump is turned off and plummets again the next time the pump runs. I think that a smoother response would be more efficient, with the solar heat being drawn steadily into the tank rather than "boom and bust" cycles of rapid pumping and zero-flow heating, and this would be more likely to happen with a more relaxed pump. I'm aiming to build or buy a PWM pump speed controller but haven't found the ideal (stingy) solution yet. If anyone has any suggestions about this I'd be very grateful I was pretty downbeat about my setup a few weeks ago because it didn't seem to be living up to its initial promise but I'm really excited now I can see what's going on. My hat is off to anyone who's set up a working and productive solar heating system without temperature monitoring, because it seems to be balanced on a knife-edge. Seemingly small inefficiencies in the system can completely eliminate the gains, or at least they did for me.
|
|
|
Post by aerojon on Sept 18, 2011 15:51:04 GMT -5
While I'm at it, I've got a little tip regarding the "Little Brown Pump" that Cye sells.A great pump and with exceptional after sales service, but in my experience a bit of a tricky thing to seal without breaking the manifold. I tried the recommended rolled PTFE tape in the threads technique but could never stop the gentle weeping from the joint. My sister's a plumber / heating engineer and recommended a couple of approaches. One thing you can try, if you're not that bothered about undoing the joint, is LS-X Leak Sealer (http://www.screwfix.com/p/fernox-ls-x-leak-sealer/23614). It's a silione-based thread sealer. The other thing is a fibre washer - really simple, basic plumbing technology, but if you've got two flat mating surfaces it might save your life like it saved mine. I got a pack of 100 from Screwfix but the product code I've got doesn't seem to exist any more. These look similar though www.screwfix.com/p/fibre-washer-pack-of-20/29825I used both of the above so I'm not sure which was the real winner, but I've got a perfect pressure-tight seal with very little torque applied to tighten the joint. As I said above, I broke my first pump head after several unsuccessful assemblies (generously replaced for free by Cye) so it was a big relief to have a working example without fretting about tightening it just enough but not too much.
|
|
|
Post by aerojon on Sept 19, 2011 4:39:49 GMT -5
Here's an updated schematic with the new pump position, and showing the locations of the temperature sensors (the two on the roof are in my head - haven't fitted them yet!) Attachments:
|
|
|
Post by cye on Sept 20, 2011 16:38:12 GMT -5
Thanks for the update Jon and the tips, I think most people with solar water panels have much smaller tanks than your thermal store and it's more obvious with a smaller tank whether the system is working.
i agree that a large tank such as your's needs monitoring.
Panel heat absorption efficiency is greatest when the panel is kept cool, so a very low pump speed will not necessarily be that efficient if it effectively maintains a high panel temperature. high panel temps being particularly inefficient on a flat panel compared with vac tube. minimising pump speed would therefore normally not be that important unless you were pumping directly to a tank , i.e., without a heat exchanger, but i now see you have a heat exchanger coil, so maybe i gave you a bum steer before.
if you're still keen to experiment with lowering the pump speed please PM me - i may have a spare PWM controller somewhere which you would be welcome to experiment with, provided you let us know how you get along!lower pump speed will however result in lower head, and there will eventually be a pump speed below which the pump will not overcome the flow resistance offered by the narrow bore pipe. DPmiller used a clever technique of feeding a small pv panel alongside a psu to power the pump, and this provided an increased speed when the sun was brighter - quite ingenious really (see his post for details)!
|
|
|
Post by aerojon on Sept 21, 2011 9:33:48 GMT -5
[PM sent]
Point taken about smaller tanks - I love my huge tank but it's a bit of a beast and obviously takes a long time to respond to input.
I read dpmiller's threads - nice idea with the PSU-PV-diode arrangement. I had considered a variable pump speed, rather than just constantly lower, but hadn't got very far with it. This is an elegantly simple solution.
Now I've written my own temp monitoring and logging system and I've got data piling up in the database I'm slightly veering towards computer control of solar motor (and wood boiler). I'm a reasonably competent programmer and a terrible electronic engineer so I might have more success playing to my strengths.
It's all still up in the air though - there are lots of things to try in search of the elusive goal of a tonne of water getting hotter for free!
My next step is very old-school: insulation. I slit all my Armaflex insulation to get it on and have been securing it with cable ties. I've got the adhesive now so I can make it properly air-tight. Got some more insulation as well so I can spend a few happy hours sticking scraps into all the little gaps around pumps, valves etc, and maybe double-insulate the most important pipe runs.
The fact that my tank loses heat faster than the specs suggest must mean that I've got insulation issues elsewhere... or the manufacturers were being "optimistic" in their rating. In any case, more insulation is never a bad idea.
I'll be buying some of this roof insulation (http://www.actis-isolation.com/produits.php?p=3&l=3&rub=41&gamme=0) soon, for my kitchen roof, and I've started wondering if wrapping the tank in a layer of it would make sense. It's about £9 per square metre so I could probably wrap the tank for under £100. Hmm....
|
|