Thermo-Syphon

This kind of set-up is also sometimes described as a "sidearm boiler." The combustion appliance (furnace, stove, etc.) contains a heating coil arranged for continuous upward water/vapor flow. Water goes by gravity from the bottom part of the tank into the bottom of the sidearm coil. Part of the water vaporizes, and a froth of vapor and water exits the top of the coil and goes to the top of the tank. The column of water/vapor is less dense than the water alone, so gravity continues to push all water into the coil.

A potential hazard would be that if no hot water is drawn off for use, all the water could be turned into steam. To prevent this, the sidearm coil can be located away from the hottest parts of the combustion chamber. If the tank is uninsulated, this constantly condenses some of the vapor back to liquid. And there should be a standard water heater temperature-pressure relief valve on the tank, with discharge piped to a safe place.

A few years ago I worked in an old cannery where the cabins had this kind of system. It never had any trouble while I was there, but I imagine a sidearm coil could burn out rarely, making a mess in the stove. This was in Alaska, so the heat from the uninsulated tank was generally useful.

The storage tank could then be below the floor, heating the floor from below? The relief valve would then be just above the the coil? (vented to a safe place, of course.) I haul water, so the next obvious question is, how much water is lost when the valve lets go? And then, could it be condensed for return to the cold water side?

The top connection from the sidearm coil can't go downward into the storage tank. The water level in the tank also must be higher than the sidearm coil (or at least higher than most of it), so if the water tank is on a lower level, the furnace/stove would need to be down there too.

(There might be some tricks with "U' and inverted "U" loops that could get around this.)

The relief valve should operate only under some pretty unusual conditions, so not likely to be a big problem.

"The water level in the tank also must be higher than the sidearm coil" Is this because it works by convection (heat rises - cold sinks).

In rough terms, yes, but I don't think "heat rises, cold sinks" is very precise. "Less dense fluids rise, denser fluids sink, provided there isn't much restriction to such movement" might be more accurate.

This could be a matter of word preference, but I think "thermosiphon" implies change of state of the circulating fluid; whereas "convection" doesn't necessarily involve change of state.

I would check with Mother Earth News for hot water heater designs. Keeping your hot water from being rusty is the trick.

If you're building this from scratch, consider using a rust-proof heat exchanger in your fire box feeding up to a conventional hot water heater tank. Plumb the feed line from the bottom of the tank to the heat exchanger, and the outlet from the exchanger to the top of the tank. Supply water would be piped in low, and a hot tap draws off near the top. If you use an electric heater tank, you may be able to adapt your outlet from the upper element hole.

As it is heated, the water will circulate naturally. As long as the tank is not filled all the way up, it should not need a safety valve, just leave a top vent open, it will add moisture to the cabin.

what has changed? a very large pot with a spigot on the stove or on a block. it fills with water and you fill a smaller pail and do what you want.

We use thermosyphons all over the Canadian arctic. We use them under cement slabs to remove heat from the slab and then radiate it to the atmosphere to keep buildings from melting into the permafrost.

Here they are filled with CO2. Here is a picture showing the radiator portion of an installation.

We found an old wood stove with a hot water jacket when we moved into a shed. To use the stove, we either had to block up the water jacket, or plumb it in, so chose the latter. Water supply was from a tank with 30m head. I obtained a disused electric water tank (140L) for storage. The tank had a 3/4" cold inlet at the bottom, a 3/4" hot outlet at the top, and a pressure relief valve (1/2" port) I did not want to put it on the roof, (though that may have worked more efficiently, not sure), and actually just had it on the ground beside the stove.

After much fiddling with pipe connections, this is what worked: using 3/4inch copper pipe. The cold supply to the system came from a T off the supply from the 'mains supply' through a pressure reducing valve. That line then ran to a T between the cold tank inlet (ie the bottom of the tank) and the cold inlet to the stove. The hot outlet from the stove ran upwards (for thermosyphon) to the tank. Here is the crucial bit. The hot outlet from the stove had to connect NOT with the hot outlet of the hot water system, BUT instead plugged into the hole designed for the pressure relief valve. Then I ran an open vent off the same connection (another T) as a 1/2" pipe running up as high as I could above the roof. The height of the vent set the max pressure in the system, controlled by the Pressure Reducing Valve (as it all heated up, I could wind down the pressure to stop water shooting out the vent).

The hot water outlet from the tank then ran to the shower, the sink etc.

The crucial thing was to connect the hot water from the stove to the hole that communicated with the body of the tank (the pressure relief valve port) , and not directly with the coil (via the tank's hot water outlet) so the stove would heat the whole tank, and not just the tiny bit of water in the coil. Until I realised this, we had very hot water for about 30 seconds (I originally had the hot water running from a T taken off the line between the stove and the tank's hot water outlet. Once the coil went cold, the water running through the stove did not have time to heat up before it was sent to the shower)

Once fixed, it worked very nicely until we moved into the house. Thats another story.