SWH prototype project part 6 - prototype conclusions

Here's a bit of a belated update to where we're up to with the prototype panel. For the last few weeks it's been sitting out in the orchard soaking up the rays, and using the energy from said rays to warm up its little tank of water. We set it up on some legs, cunningly constructed from a pair of old doorframes nailed together. It's now sitting at an angle of about 50° to the horizontal*.



















(*For the best all-year-round collection, we would set the panels at an angle equal to our latitude - here it's 34.26°. Setting it up a bit higher favours winter collection (by facing the lower winter sun better), and slightly reduces the chances of the panels overheating in summer. I did lots of complicated calculations to try and figure out the best angle, but these remained somewhat inconclusive - so I reverted to a rule of thumb suggested by some people who seemed to know what they were doing...)

At present there's no pipework in place to enable one to actually make use of this warm water, but that doesn't mean this setup isn't useful. Running a closed system like this allows us to calculate, to a ball-park figure at least, how well it's performing. By measuring the tank temperatures at the start and end of the day we can calculate fairly accurately how much useful energy we're collecting (obviously we have no way of measuring the energy losses, but we can estimate efficiency by comparing our daily energy collection figures with solar insolation data collected over a couple of decades at a weather station not far away).

























Here are some numbers. If you have some objection to numbers, feel free to skip to the pictures and chatty bits lower down!

Some simple thermodynamics also tells us that the hotter the system is running, the higher the heat losses, and therefore the lower its efficiency. Which just means that a record temperature in our tank doesn't necessarily equate to the greatest amount of energy being harvested.

The greatest daily (useful) energy collection we've witnessed so far was 10.6MJ, with our 59l tank rising from 17°C in the morning to 61°C in the late afternoon. Comparing to our insolation figures for January, (9.4MJ/m² per day x 1.79m² = 16.8MJ per day), this suggests an efficiency value in the region of 63%**.

The highest tank temperature we've recorded, after it had been sitting out for a few successive clear days, was a mighty 75°C (with the panel sitting at 81.5°C, and water entering the tank at 78°C). But since the water was already sitting at nearly 50°C from the previous day, this equates to a much less impressive 5.3MJ, and 32% efficiency.


















(**All these calculations contain hefty doses of approximation and a little guesswork. We take several measurements in the tank, as the water tends to stratify in layers of different temperature. You can stick your hand in and feel the temperature lower significantly as you move it downwards in the tank. Also, the weather station's figures are averages over the whole month, for many years. If we happened to get an above average clear sunny day, then this would tend to push our efficiency figure upwards. So we might surmise that we're operating somewhere in the range of 50% - 65%.

Also, the system in it's current state could be improved significantly by the simple expedient of insulating the connecting tubes, by insulating the tank better, and by not opening it up every hour to stick a thermometer in!)

Bear in mind that these measurements are all for the middle of winter - in midsummer we would expect to gather at least twice as much energy.

In case all that maths really wasn't your thing, here's a few of the conclusions I've been coming to about this, and therefore what I would keep or change for the design of the large system:

• The basic design seems to be working well, and we have suitable solar resource and a suitable location and use for it in the hospital.
• The 1.2mm aluminium is seriously hard to work with. Find some thinner stuff.

• Get someone who knows what they're doing to do the copper soldering. Q'bad's finest looks pretty, but I wouldn't want to guarantee its quality.

• Two layers of glass is really heavy, not to mention prone to breakage. Try to get hold of some of that polycarbonate stuff they use for greenhouses.
• Don't let the panels sit outside in the sun with no water in them - it hit 103°C the other day - I suspect the polystyrene insulation in the back of the panel melted a bit. Find some insulation that has a higher melting point. Work out how to shade the panels in the event that they're sitting out empty.
• Don't use whatever this material is for lining a tank with!

It appears to degrade over a few days in the hot water, leaving this interesting greenish flaky stuff:

• Don't drop the thermometer into the tank!

SWH prototype project part 5 - temperature logging

A quick foray into some electronics, just for a change. The plan is to build a little data logger which will record the temperatures at various points in the panels and tank over the course of the day (or week, or month for that matter) - and thus save me having to take the lid off the tank and check a glass thermometer every hour!

This nifty little screen will eventually give an up-to-the-minute temperature display.



















For my fellow geeks, the red board is a Seeeduino Stalker microcontroller - an Arduino variant with onboard real-time clock and microSD slot for data logging purposes. I'll be connecting up a few MCP9700A thermistors in the next couple of days. In it's current state I have 6 inputs to play with - possibly expandable with a multiplexer at some later stage if I feel like it.

Time and part-availability* permitting I might think about adding a light sensor of some sort, and an alarm for if/when the panels get too hot in summer (so that someone can take some sort of action - probably just shading the panels. It'd be sweet to have some sort of canopy unfurl automatically, but I think that's probably a bit over-ambitious!)

(*a big problem - trade embargoes at the UK end mean RS orders take 6-8 weeks to arrive, after they've obtained an export certificate - I had to get our recent visitors to bring out the thermistors for me. On the other hand, the circuit boards arrived from Singapore without too much hassle...)



















The absence of breadboards means a whole lot of trying to solder ends of wires to each other - a bit messier than I'd like.

I think this soldering iron is designed for somewhat larger applications...

SWH prototype project part 4 - a bigger tank

See also previous instalments.

A solution to the leaking tank problem:



















We lined the box with some plastic sheeting* and after a few unsuccessful attempts found a way of connecting the tubes in a non-leaking manner using a short length of metal pipe and those metal hose ties you tighten up with a screwdriver. But then it was time to think a little bigger.

(*One of the recurring issues we have here is the question "what is this made of?". Will it melt? Dunno. Will it degrade over time? No idea. Will dangerous chemicals leach into the water? Who can tell? (Don't worry, this is only a temporary solution, and no-one will be drinking the water!) At the bazaar WYSIWYG applies, but that's only any use for those properties measured using the eyes and/or fingertips.)

Also, behold Peter, engineering student from NZ who is over for 6 weeks. In that Christian community small world way, it turns out he knows one of our friends from home...

And now behold one of our workshop buddies, ignoring all health-and-safety notions in true Pakistani style. While this photo doesn't belong here in any kind of thematic sense, it does in a camera-memory-card-chronological sense, and it looked cool. So there.



















So here's baby tank's big brother in situ, constructed in a similar manner (wood frame supporting polystyrene insulation lined with plastic sheeting). This one contains about 59 litres, including the water in the panel and tubes. As you can see, in its current position, the sun disappears behind a tree around 2.30pm, so we've got a bit of repositioning to do.

And here is a pretty rainbow-coloured graph depicting the other day's temperature variations - this tank got up to around 45°C, which is a fair bit lower than the small tank (as you would expect), but actually represents nearly a 3-fold increase in the amount of useful energy harvested. (Lower tank temp means lower heat losses, and this tank has a relatively smaller surface area to lose heat from.)

Towards the end of the day we got round to drilling a hole in the top of the panel so we could measure the panel temperature, hence the blue line appearing at the right hand side.



















To be continued...

SWH prototype project part 3 - it lives!

Work on the solar water heater felt like it had ground to a halt somewhat this week due to one thing or another, but yesterday we finally got it moved out of the workshop and into the sun on the other side of the compound. (This was no small operation, our ridiculously overweight panel taking 5 men to lift, with considerable effort - we're going to have to rethink the idea of putting it on the laundry roof!)

This is very much a "lets-just-see-if-it-works"-type test, and thus has all sorts of caveats and problems built in from the word go. The first one being that as soon as the plastic piping we used to connect it up warmed up slightly, its flexibility increased, and it sagged, immediately preventing the thermosyphon effect from taking place.

(A quick lesson: In order for the water to circulate by itself, the piping must be 'uphill' all the way from the lowest point of the loop to the highest. Then, as the water is heated in the panel, it becomes less dense and rises up, whilst the denser, colder water in the tank sinks down the other side of the loop into the bottom of the panel. The water heats up a little more on each cycle through the panel, gradually raising the temperature of the tank water. The neat thing about this setup is that at night the circulation stops naturally once the panel temperature drops below the tank temperature, thus preventing the panel from radiating away all that heat you've carefully collected - and no valves or pumps required.)



















Problem 1 solved (for the time being at least):



















A most encouraging sight (hot water entering the tank - click on the picture for a larger, clearer view):

























Problem number two. This impromptu water tank is made from a polystyrene cold box I found lying around. I knew it was going to leak, but it managed nearly a day-and-a-half before it became a problem - more than enough for the first try:



















It's also a very small volume for this panel, around 13 litres including the water in the pipework. The next step will be to make a tank bigger than this, yet smaller than the big blue behemoth - perhaps around 50 litres, with decent insulation, so we can gather some more useful data from it.

And on that subject, oh for a nice datalogger and some thermistors! At the moment I have 3 glass/liquid thermometers of slightly dubious accuracy, which I obviously have to directly observe every time I want a reading.

Still, we hit 63°C after a couple of hours, which is nice. It was too hot for me to rescue my thermometer by hand when I accidentally dropped it into the tank.

Projects for the next few weeks:

• make a ~50 litre insulated tank.
• move panel another 20 yards so it's not shaded at certain times in the afternoon.
• give it some legs to prop it up at the right angle so it's actually pointing at the sun.
• build a platform for the water tank to sit on.
• connect it up with some proper pipework.
• run some day-long and week-long tests.
• work out how on earth to strip about 50% of the weight out of the design!

SWH prototype project part 2

Yesterday I finally got the panel connected up and heating some water, but before I recount that story in the next post, I thought I'd better fill you in on the rest of the panel-building. Here's the panel innards more-or-less complete, after much sledghammering and much nut-and-bolting:



















And now painted black and dropped into the outer frame:



















Our spray-paint claims to be flat black, but it does have a distinctly satin finish to it. Better than straight gloss, but I think we'll shop around for a bit longer next time.

Two layers of glazing in place, and in the process of being secured down:



















We sliced up some old tyre innertubes for use as gasket material. 6" x 4" sheets of glass are not easy to work with, and contribute to this thing weighing far too much. We managed to get both sheets in with only one small corner damaged, however.

Next, we 'reclaimed' this unused water tank from the roof of the operating theatre:



















This one is actually far too huge for our prototype panel, and will be more suited for when we have the much bigger array of panels for the laundry. Perhaps we'll be able to run it at quarter volume or so for now.

Because it'll (hopefully) be holding warm water, we have to get some insulation around it:





















This is a start using what we had lying around, but I think we'll have to make a few improvements to it before we use it for real. All this wood has been used at least once before - removing all the old nails and so on adds at least the same amount of time and effort again as the actual construction does.

On to the testing...

SWH prototype project part 1

Last week my solar water heater project really got moving. Here's a quick photo tour of what we've done so far.

Job no.1 - make the planks.



















Completed outer frame. That's the easy part over with then.



















Then lots of cutting up of bits of copper tube. The white board is used in conjunction with a steel rod and a great big sledgehammer to form fins out of sheet aluminium which will fit around the copper tubes.



















We took all our bits of tube out to a chap in the bazaar who has an oxy-acetylene torch to get it soldered together. As in everything here, health and safety is a bit take-it-or-leave-it: there's a pair of dark glasses the guys sometimes shove on while they're arc-welding, but that's about it. This is our first attempt at leak testing, which wasn't entirely a success (well, the test achieved it's aim fine, but you know what I mean!). Cue a return trip to soldering man.

























I don't know much about soldering copper, but I wouldn't like to hold up our pipework as an example of the greatest quality workmanship! It does look rather beautiful though...



















Here we are starting to lay out and fix the fins over our copper opposite-of-a-radiator arrangement. The fins, having been formed by much bashing, aren't too bad a fit to the tubes, but probably nothing like as close as any of the efforts on the DIY sites I've looked at. We run lines of silicone gel next to the tubes to try and fill up any air gaps (while not a stupendously wonderful heat conductor, it's a darn sight better than an air gap; also, it should be able to cope with high temperatures and differences in thermal expansion of the different metals).

























The aluminium sheet we were able to find in Pindi is 2 or 3 times as thick as what I would have liked - at around 1.2mm or 1/20th of an inch (getting used to everything in imperial is a bit of a challenge too!) - you can bend it in your hands with a bit of effort, but getting it into a shape that'll fit snugly around the tubes was not easy. This added thickness caused all sorts of problems, as compared to the reasonably effortless-looking examples I'd been reading up on:

1. Cutting all that sheet up with only a pair of tin snips is hard work. Really hard work.

2. Shaping the fins took 5 mins or so of bashing about per fin (DIY site example time: 15 seconds).

3. Other people's fins are splendid, shiny, perfectly-formed works of art. Ours look like they've been battered with sledgehammer. Oh, wait...

4. Other people lay out all the fins and gratuitously attack them with a staple gun. Us: staples won't go through, nails won't squeeze them together, split rivets won't reach through, the pop-riveter is broken. So it's out with the drill (of course, plugged into a dodgy extension lead of bits of wire twisted together and shoved in the socket), and lots and lots of nuts and bolts. This has the benefit of pulling the fins into close contact with the tubes, but has the decided disadvantage that we have to lift the whole board up to tighten the nuts up. With a screwdriver and a pair of pliers.

A couple of days hard graft, a couple of hundred holes, and a couple of hundred two-man tightening-up-of-nuts-and-bolts later:

























The last batch of little fins drying in the sun (a little coat of paint on the contact surfaces should help prevent galvanic corrosion) - these ones will fit over the thicker manifold pipes at the end. Then, hopefully, it should be all ready for some black spray-paint.