Unless you use evacuated tube solar water heaters (such as the Apricus, Mazdon, or Solarmax types), I wouldn't want to leave water in overnight. That'd be an awful lot of heat loss.

You didn't mention anything about a siphon before. I was picturing a closed loop with a pump. At night, you'd shut off the pump, one side would drain out, the other would suck air, and the system would be empty. Were you picturing passive circulation or something?

Getting water to boiling at 1atm without evacuated tube heaters shouldn't be easy, even in Texas (I know evacuated tube systems can pull it off, though). Whether boiling would be a problem depends on how your system is plumbed. If it's pressuretight, and you have a sudden addition of pressure, obviously that's a potential recipe for rupture. Potentially catastrophic. It all depends on your system. Never hurts to have a pressure release...

Karen, thanks for the response.
I was invisioning a passive thermal siphon without a pump.
I can't afford fancy evacuated solar panels; they will probably be homemade. A pressure release would be a good idea.

If it was a passive siphon I was thinking I would just have a valve to shut it down either during the day for summer or at night for winter. With the panel on top of the roof maybe a passive siphon wouldn't work . . . not sure. Maybe I should make it ground mounted and movable. I will have to think about this some more.

By a closed loop do you mean that both ends of the pipe would end in the cistern or do you mean truly closed like sealed?

K-

What I was picturing (which seems to be different from what you were picturing) was an active siphon system that involved a cistern with an airspace above the water. The "cold water out of the cistern" pipe reaches into the cistern at a level in which it will be submerged at all times, and the "warm water into the cistern" pipe only reaches into the cistern at a height that will always be above the water level. The pump is connected to the "cold water out of the cistern" pipe and pushes the water from the cistern through the solar heater; it eventually makes it through the heating loop and falls back into the cistern, warm. When you shut off the pump, you have a scenario where you have a pipe that's full of water, with one end lower than the other, and air to be sucked into the higher end -- i.e., a siphon. The water will go backwards through the pump (assuming it's a type of pump for which this is possible!) and back into the cistern.

This is just what I pictured. But there are many possible designs. Examples: You could lighten the load on the pump by having the in and out heights be the same (both submerged) if you add a valve to let the system suck air (which you open when you shut off the pump). You could use a pump that doesn't let water flow backwards through it when off if you add in a valve to open up a bypass around the pump, also to open when you shut the pump off. Also, you can use an "indirect" heating system by pumping glycol instead of water, and using a heat exchanger, thus preventing the need to drain the heater at night. Your heat exchanger could be little more than a coil of tubing in your cistern.

However, these aren't what you're picturing. If I understand you right, you're picturing what's called a thermo-passive water heating system. You're right to be concerned about altitudes -- a thermo-passive system has to have the cold water *above* the heater, because the hot water wants to rise. Also, fundamental to any siphon is that if you ever drain it, you need to refill it, either with a pump or manually. Also, passive siphons have a much lower flow rate than a pump-driven system. The higher the cistern is over the heater, the faster the water circulates, but it's hard to get it up to the rate of an active system.

The basic design karen suggests is called 'drain back' and is the safest design. It will not be damaged by boiling as the system operates at (and even below) atmospheric pressure. Nor will it be affected by freezing, as all the water will drain out long before it gets that cold. If the outlet is nearly at tank level the system will only need the energy to lift the water the different between the water level and the outlet level, plus the pipe friction - this can be very little energy. The trick is to use a pump that can prime the loop itself (some people use two pumps, one switched off when the loop fills).

I think conventional panels are too expensive for heating thermal storages and would suggest a much simpler system using water trickling down some old roof iron covered with PE greenhouse film and perhaps a layer of polycarbonate. This can be part of a shed or even mounted on your house roof. A good pond pump can lift water say 5m at a flow rate of 10l/min, this means you can transfer about 700W of heat per degree C. If your trickle heater adds 10C that's 7kW of heat, requiring a collector area of at least 7m^2 (more likely 20m^2 - perhaps a rectangle of 4m*5m). 7kW would heat my 4.5kL tank 1 degree every 45 minutes.

(And because the pump is sitting in your thermal storage, it won't get cooked - you'll struggle to heat that amount of water much above 30C due to losses; especially if open to the air)

I was picturing a thermo-passive system but I think the solar panel should be mounted on the roof because it will be hard to manage if I have to move it around on the ground. I didn't think it would take much energy to move the water but was hoping to go with a passive system because it simpler and requires less maintenance.

I tend to overdesign and then my partner comes along and says wouldn't it be simpler to just "blah". He is usually right so I was trying not to fall into that trap. I have had good luck with my pond pumps not failing but I have had several smaller ones do so. Its just another moving part that can go wacko on you. It sounds like a neccesary one though. If I decided to use the glycol with heat transfer to the cistern how would that effect efficiency? Do they break down / spring leaks easily?

Nathan, I am not sure why you think that regular panels would be too expensive for heating thermal storage. I have always wanted to build one. Are you just thinking of the sheer size? My tendency would be to put one up that was say 2X1m see how that performs and then add another if neccesary. I know the copper can be pricey but a lot of the other stuff I have laying around. I pretty much would just need the copper and the glazing. Since we have big hail I was thinking of using that polycarbonate 3 wall stuff. My Dad said it was really tough.

The other system is simpler but in my climate I always have to worry about evaporative loss. Also even if I could seal it I am not sure my partner would go for continuous flow over the roof.

Nathan you are working with 450 l of water in your thermal storage tank right? Thats about 118 gallons (please forgive me using this antequated system but one of our brilliant presidents nixed the conversion project when I was like 12). Anyway thats about 2 55 gallon drums. I am figuring a minimum twice that volume and maximum maybe 385 gallons or which would be about 1450 liters. I would need to go back to my college chemistry text to do all the figuring.

I know thats a lot of heat though. . . For my purposes though I think if the collector could just keep up with loss in the winter so I could maintain a reasonable air temperature I would be pretty satisfied. It sounds like my best bet is to install it in the summer so I can get all that water warmed up before winter hits.

Another question. If I decided to use a backup heater in the winter would it be more efficient to heat the air or the water? Maybe not as efficient to heat the water but certainly it could store more warmth . . .

I think conventional panels are too expensive for heating thermal storages and would suggest a much simpler system using water trickling down some old roof iron covered with PE greenhouse film and perhaps a layer of polycarbonate.

Clever. That'd be amazingly simple and cheap to build. And tempting. Instead of iron, I could use the same pitch-black shingles used on my shed, along with two layers of my surplus PE sheeting. A couple two by fours to prop it up, a dripper hose on the top, regular hoses to connect, a pond pump, another roll of aluminized insulation, perhaps two trash cans for addl. water storage...

Darn you for tempting me! Must... resist... I need to avoid working more on my GH until I'm ready to commit to build the larger scale one I've been designing. ;) Which means waiting until my partner either gets a job here or decides that we have to move.

The problem is that at best you're getting maybe 800W/m^2 in winter, and you need lots of watts. 10m^2 of copper is going to be tedious to handle, expensive and environmentally unsustainable. Especially when you could use the existing roof, plus maybe a $1/m^2 to do your heating. a 2x1m panel might net you 43MJ on a sunny winter day (heating your tank by 2C). You probably want 10 times this in practice.

The big difference between DHW panels and solar thermal panels is that you a) don't need high pressure, b) you don't need high efficiency (instead you are optimising $/W) c) you are not in the business of selling small and expensive boxes to be shipped around the place, instead you can handle bulk materials easily. d) the temperature differential between the water and outside is relatively low, meaning simpler insulation.

It is worth learning to use metric (SI) units, as many calculations become easier. In the metric system 'k' means 1000, so 4.5kL means 4.5 * 1000L = 4500L = 1190gallons.

1L of water stores 4.2kJ / kg (and 1L weighs 1kg) degree C.

So my 4.5kL tank will store about 18MJ per degree C change.

For cooling the radiator is the way to go IMNSHO, for heating I would use underbed tubing as mentioned by cuesta in another thread. I am currently not using any valves - instead all the water circuits run all the time, I instead modulate the fan speed. Under bed warmers are much more efficient, so I might keep the soil at 25C for 24 hours if my tank is 26C. Thus I can collect a lot of heat during the day and release it slowly at night. But doing this properly means knowing whether it is more important to stockpile heat and release it slowly at night, or stockpile cool and use that during the day - hence computer and forecast.

It would be more efficient to heat the water, as you can then direct the water's heat where it is needed. Also, you can then consider heating the water with high burn, high efficiency, renewables such as local wood. Alternatively, you might use a compost pile heater with some water loops. Also, heating the water means you can simplify your control system (the water heater kicks in when the water cools below 15C or something).

(The problem with writing a long screed in the spare minutes of the day is that people post while I'm still halfway through :)

karen: I would love to use asphalt tiles or paper, but it is not available in australia. My wife suggests it might not hold up in the high temperatures inside a collector if you turn the water off (such as during summer). I reason I suggest gal-iron corrugated roofing is that it is easily available here (indeed I have 40m^2 of it stacked in my yard already) and zinc is an excellent selective surface (which is why it gets so darned hot). A huge advantage of asphalt tiles is if they are the ones with the little rocks they will cause the water to spread out evenly over the whole surface maximising the efficiency and reducing the amount of effort in balancing the flows.

Another issue is that zinc passively discourages biology, my rockface experiment shows that water + sunlight = cyanobacteria. These are usually black, and are excellent filters for the water, and also sequester CO2. They might improve efficiency, or they might tend to block the water flow. I suspect they will be replaced with mosses and ferns after a while, which will dramatically reduce the collector efficiency (but look nice in return :).

Apologize for what may be a stupid question, but I'm following this and the other thread and learning a lot.

I think conventional panels are too expensive for heating thermal storages and would suggest a much simpler system using water trickling down some old roof iron covered with PE greenhouse film and perhaps a layer of polycarbonate. This can be part of a shed or even mounted on your house roof.

Am I picturing this right...you have a thin solar trap which the water runs through? Like a flat, water-filled mini-greenhouse on the roof?

stress: precisely. Very cheap, large surface area, well known technology. thin = 5cm from film to roof or less; cheap = cost of greenhouse film + structure (perhaps wood with screw-down battons or something). Not completely flat, as you want the water to drain out at night or whatever. (although in fact such as system would not be affected by freezing as the ice can expand upwards).

Indeed, at night, if you use a non-IR block film you can actually cool the water below ambient by radiating out to space. Our house has a roof that slopes north and south, I have been contemplating building such a collector on both sides of the roof, using the south facing side for cooling and the north facing for heating. A shaded wall facing the sky can cool by radiation even during the day - recently it was 40C ambient, yet the sky was -20C: by stefan-boltzmann equation this could give 230W/m^2 of cooling power. (compare at night giving 360W/m^2)

But then in that case you are talking about heating/cooling the structure rather than storing heat for the GH, is that right?

stress: I don't understand the question.

In the light of a new day, I see that it was a ill-worded question and after rereading your last post this morning, I got it.

Sorry for the threadjack, Kate!

Well if I could make the trickle through collector neat looking enough I could probably get it by my partner.

So still working on the logistics here. You are talking about putting some sort of UV resistant waterproofing (I would rather have the whole structure be fairly low maintenance) on top of the roof surface in a sheet, putting the galvanized iron on top of that as the heat collector, and then topping it with some sort of film to trap the heat in. You have a pond pump going all the time that you want to be trapping or radiating heat. The water trickles to the base into a pipe / gutter and back into the thermal storage (in my case the cistern). So there is not as much temperature differential and the system depends on high volume recirculation to make up for that. I guess the slope of the collector would be pretty critical because of how fast the water would drain off of it (a greater slope would me less temperature differential).

Do you think you could make it fairly water tight? Are you at all concerned about leakage onto the roof?? I am not sure asphalt shingles like to have water trapped against them.

For the cooling how could you possibly get the temperature of the water below ambient without using evaporative cooling?? Or did you mean the ambient temperature of the thermal storage medium? Also how can the air temperature be 40C with the "sky" temperature at -20C? I am sure I am missing something here. Please explain.

Not to worry stress I always learn from your sort of jacking.

Kate:

I'll address the "sky temperature" issue. You're probably familiar with the concept that all objects radiate energy. At temperatures we humans are used to, this is typically in the mid infrared range -- 8-12 microns wavelength. An "ideal" radiator is called a "blackbody". "Emissivity" is an approximation of how far you are away from being an ideal blackbody. The amount of energy lost is the Stefan-Boltzman constant * emissivity * Area * Temperature to the fourth power.

A person has about 2 square meters of surface area and an emissivity near 1, so that's 5.67e-8 * 1 * 2 * 305 ^ 4 = 980 watts. That's an awful lot of energy -- tens of thousands of "food calories" per day. Thankfully, while you're emitting energy, you're constantly receiving it from everything else that emits it. A room temperature (20C) body will be emitting about 835 watts in that same 2 meters squared. Factor in clothes absorbing and re-emitting your radiation back to you, and gets down to about 2000 food calories per day. Now, this gets complicated, as you're not perfectly surrounded with objects hovering just over your skin radiating room temperature-emitted IR back at you; you have to deal with complex geometry of your skin and the things around you. But you get the basic idea of how heat exchange through infrared works.

Now, not only are things on the surface radiating, but things above you are as well. However, the atmosphere isn't a nearby solid. You're getting radiation from a broad range of particles, from ones radiating right above you, further, further, and further away, all the way out to the frigid edge of the atmosphere and beyond. It is the net radiation of these particles that matters in determining how much incoming infrared is absorbed by something that faces skyward.

This is one of the big reasons why clouds are insulative. When you have cloudcover, you tend to be mostly exchanging radiation with the lowest layer of clouds, not with a much broader range of air particles (including very high altitude ones) and the cold of space (which is about 2 degrees kelvin due to the cosmic microwave background... this gets complicated). The water droplets absorb IR well, and reemit radiation that came from the surface that normally would have been directly lost to space. Of course, having cloudcover during the daytime cools because they reflect away the visible light from the sun ;)

karen has given a good explanation of sky cooling.

regarding the roof: You will have to think carefully about the water issue - tiles do not stop water getting in, e.g. A sheet metal roof on the other hand is exactly the same material they use to make water tanks here:http://www.bluescopesteel.com.au/go/product/aquaplate-steel-for-water-tanks

And as long as the screws are all done with rubber grommets on the ridges you should not have any trouble with water penetration.

Practically you will have no trouble with the volume - water holds a danged lot of heat so a 10L/min trickle will carry away 700W for every degree C you heat it up. My pump is rated to 50l/min.

It is worth trying to minimize air leakage as evaporating water will take even more heat away (aside: you could design your system to evaporate water and recondense it in your tank, dropping the water rate to a measly 175ml/min)

Remember the aim is to make the most watts per money spent. If going from 25% efficiency to 50% efficiency is going to more than double your cost per area it's probably not worth while. Our tiny house has a roof area of 100m^2 (no doubt your roof is much bigger) - that's an average of 800MJ per day in Melbourne or 9kW of heat continuously, day and night. If our roof is already made from epoxy coated iron I might be looking at $300 to do the whole roof - 3 cents per average watt, or 0.3cents per peak watt (compare with PV at around $5/peak W)

(for comparison, 800MJ of propane would cost about $10US so such a system would pay off based on fuel alone in less than 30 days)

If you examine asphault shingles up close, you'll find that they're layered one over each other -- three deep in the case of mine. From a practical standpoint, while they won't keep standing water out, they should keep running water out just fine. I'd be more concerned with algae growth than leaks.

I know a sheet of metal roofing would hold the water away from the roof. I guess if I was careful I might be able to just use the metal and one piece of glazing. But wouldn't another piece of insulation under the metal make it more efficient? Bending the edges up just a bit might make me less paranoid about leaks. Another option as apposed to screws is to use construction adhesive. It is really tough stuff. You folks are making me itchy to get out in my shop. After X-mas maybe. I think I should probably build a very small experimental trickle through collector just to try it out and see how I do with leakage.

about the "sky cooling" . . . It all makes sense for the most part. Though biology is more my thing than physics. If you wouldn't mind explaining a bit more I would appreciate it. I am having trouble connecting the cooling of the panel with your explanation karen.

So lets say you have a panel that is circulating warmish water in it that you want to cool down. You put it on the shady side of the house on a day when the air temperature is greater than the temperature of the water you want to cool off. So the water molecules release IR radiation into the panel and the panel passes that IR on to the air. The air is made up of all sorts of stuff, oxygen, nitrogen, CO2, water vapor & then a bunch of trace stuff. So here comes our IR wave and it runs into a gas molecule. With sunlight the gas molecules would absorb some energy and start bouncing around faster. We would measure this as a higher air temperature with a thermometer. I would imagine that the with the air being warmer than the water in the panel it would do the same thing in terms of energy transfer and warm up the water.

Does IR work differently? Do some gases reflect the IR or not interact with it at all? This seems to be the key paragraph that is confusing me.

"Now, not only are things on the surface radiating, but things above you are as well. However, the atmosphere isn't a nearby solid (ARE YOU IMPLYING THAT THE ATMOSPHERE RADIATES DIFFERENTLY BECAUSE IT IS A GAS INSTEAD OF A SOLID?). You're getting radiation from a broad range of particles, from ones radiating right above you, further, further, and further away, all the way out to the frigid edge of the atmosphere and beyond. It is the net radiation of these particles that matters in determining how much incoming infrared is absorbed by something that faces skyward (SO YOU ARE SAYING THAT THE NET INCOMING RADIATION FROM THE ATMOSPHERE IS LEss THAN THE OUTGOING RADIATION FROM THE PANEL?)."

I am still a bit lost. Sorry for needing so much explanation. If you ever have frog questions feel free to ask and maybe I can reciprocate a bit.

Cooling down your heat sink.

I am a bit simple minded I admit it.

If your water in your heat sink is getting too warm then dump it!

Re-fill it with nice cold water!

(do laundry with your hot heat sink water) (flush the toilet with the hot water) you get the idea....

Use the hot water for something you would have used water for anyway, no water loss there. But you trade hot fot some cold.

OK I didn't mean it to be as simple as that. Water your outside plants with your thermal storage water and top up your tank with cool new water.

If I was in the pacific northwest that would work fine Chris. Unfortunately I live in a very arid area where our aquifer is on its last legs so water is hard to come by. I sink PVC pipe next to the larger plants I put in so the water goes directly to the root zone without too much evaporation.

The other thing is I will be using rainwater instead of the city water which is so full of dissolved solids that one year they had to issue a warning for people on a low sodium diet not to drink it. The hard water tends to evaporate and leave behind a lot of deposits that gum up toilets, faucets, valves, whatever. If it was to rain while & I had a high heat load then I would definitely go for the trade out thing.

In the short term I could use the hot water for other things but that involves either a lot of carrying buckets or a lot of plumbing.

Anyway I am just one of those weird folks who wants to know why so thats why I am asking so many obnoxious questions.

ok kate:

Extra insulation behind would indeed improve efficiency, but maybe your roof has this already. Remember, the game we're playing is $/J, not maximum efficiency. Is the insulation going to have a big effect? I suspect not, because the air behind is stagnant anyway.

The metal roofing I'm thinking about is made in corrugations about 3inchs across and is always laid so that water can't get into the roof. So its edges are already upturned.

Firstly, you would need to insulate your sky radiator from the ambient air (two layers of PE film would do this). Then you must realise that dry air is completely transparent to thermal IR (and visible light - otherwise we would be able to see through it!). So the thermal IR keeps travelling until it hits something, perhaps space. On the other hand, the sky itself, being colder, emits much less and thus the object cools. I am not convinced that the sky is truely a blackbody at -70, my IR thermometer may not be telling the entire truth.

sky cooling is of low importance anyway, as evaporative cooling will work as well. in practice (on a clear day humidity is low).

"SO YOU ARE SAYING THAT THE NET INCOMING RADIATION FROM THE ATMOSPHERE IS LESS THAN THE OUTGOING RADIATION FROM THE PANEL?" Yep.

Dumping water would not be practical here in Melbourne where we aren't even allowed to water gardens, but the idea is good. Also, it would involve dumping a lot of water! However, to run some quick numbers: to dump 4kW of heat means 1kg C/s, nearly 16 gallons per minute for a 1C delta. Whereas evaporating would require just 4floz/min.

Thanks for you patience Nathan & Karen. I get it now. I must have had a fuzzy brain yesterday. Oh well, thanks again.

kate

P.S. As I was cleaning out under the hottub deck I found a bunch of rigid foam insulation so I will probably reuse it. no $