Rocket Stove – Hot Water!

by Tim Barker

The basic unit, minus the rocket stove and piping

In my last post, I showed a picture of a wood powered water heater, so now we’ll roll up our sleeves and get into how this was designed. But first a warning! Boiling water is easy to do, but boiling water in a closed container and not blowing yourself up is much trickier — in fact I’ve heard it said that there is the equivalent of a stick of dynamite in 500 grams of boiling water! So if you blow yourself up, be it on your own head. Having said that, I have spent a fair bit of time creating a design that is simple to build, safe and efficient.

So, first let’s consider some basic principles. Contrary to what your mothers may have told you your father wasn’t so bad a cook that he could burn water, as you can’t get it over 100°C as long as there is water in the pot. The boiling water effectively dissipates the heat energy as quickly as you can put it in. If however we raise the pressure by putting the water in a container and seal it then we raise the boiling point of the water. Some of you by now are saying "congratulations, you’ve just described a pressure cooker". Nods to you, that’s exactly what I’m getting at. As every schoolchild knows, water boils at 100°C, but remember there is the caveat "at sea level". If we raise the pressure above that then the temperature at which the water boils will also rise. Correspondingly lower the pressure and water boils at a lower temperature. For instance at the top of mount Everest water boils at approximately 69°C. Okay, now hold that thought while we forge on with the water heater design.

A little context: PRI Australia, where I worked for the last two years, had a need to heat water for up to 55 people per night during their PDC courses. Obviously anything we did should showcase appropriate technology while being easy to use and live with. Solar was considered but a unit to heat that much water was both a huge undertaking and very expensive if we were to purchase it. A compost shower had been tried but issues of volume and heat transfer made it inappropriate and, worse, cold. Since my days as an intern at PRI I had been keen to try some ideas I had and so put forward an evil plan to build a wood powered hot water system. The power unit, so to speak, would be a rocket stove and this would In turn heat a drum of water positioned above it.

Now for the details, I arranged the drum (an old 180 litre stainless hot water system tank) on its side above the rocket stove, which in this instance is built of old red house bricks for the feed tube and burn tunnel, while the heat riser was an old stainless flue pipe insulated with perlite. This came up to within 50 mm of the hot water tank and towards one end.

The shell of the unit showing the 200 litre drums and the end
of the hot water tank just visible

Around the hot water tank I arranged a cylinder made of two 200 litre drums welded together with the ends cut off. This basically made a cover around the hot water tank that had a gap of approximately 50 mm all the way. This is where hot combustion gases passing out of the end of the heat riser travel around the hot water tank and give up heat to the tank. As the gases give up their heat they cool and are displaced downwards and eventually pushed out of a hole, low on the end of the outer cover. To stop heat passing out of the outer cover we then wrapped the 200 litre drums in recycled glass wool insulation and then some old corrugated iron around that to keep the rain out and everything tidy.

An end view showing the outer corrugated iron layer, the insulation layer
and the inner 200 litre drum with the end of the hot water tank
visible through the hole

So the gas path is easy. It’s the water part of the system that has to be gotten right or things could get ugly.

So let’s go back to our hot water tank. We plumb in a cold water inlet from a pressurised source (mains or tank) with a tap and we use this to fill the tank. At the top of the tank we have an outlet, so when we open the tap the water fills the tank and overflows out the outlet. We then shut off the tap, and the tank is now full of water. If you want a simple system to fill a bath or for hot water for washing clothes this would be okay, but remember that because we have no tap on the end of the outlet we rely on the overflow for the hot water, so turning the inlet tap off will not immediately shut off the overflow which will continue for a while longer. Also, if the water boils in the tank you can have hot water and steam slugging out the overflow pipe.

To get a system that supplies hot water on demand means we have to do a few more things. We take a coil of copper pipe (13mm x 18 metres long) and put it inside the hot water tank. In my case I cut a small hole in the tank and ‘wound’ the coil in. Each end of the coil then projects out each end of the hot water tank. In the tank I used there were holes already there for the old connections so I passed the pipe through and silver soldered it in place and then welded back in the bit I had cut out to get the coil in. If you don’t have welding skills, don’t despair — all this work can be done at your local welding shop and should be comparatively quick and cheap. Another variation is to use an old clip-top 200 litre drum, sitting upright, and pass all the fittings through the lid and use compression fittings, so no welding — but be warned, the drum will rust out eventually (probably two or three years).

Okay, let’s recap. We have a drum, it has a pressurised pipe leading into it with a tap, we have an outlet or overflow, we open the tap and the drum will fill and overflow out the outlet. We have a copper pipe inside the tank, and where the copper pipe goes through the walls of the tank it is soldered in place so water in the tank will not leak out. The water in the copper pipe does not mix with the water in the tank.

Now we connect the copper pipe to the drum fill pipe, but we do this before the tap. The other end of the copper pipe then goes to your end use, be it shower or sink or whatever, and there is a tap on that end. This means when we open the shower hot tap, hey presto, water automatically flows.

Now is the time to remember our pressure cooker example. The water In the tank is not pressurised or able to be pressurised because it’s open at the overflow. It will boil at 100°C at sea level. The water in the copper pipe comes very near to the temperature of the water in the tank because copper is an excellent conductor and because of the length of the copper pipe, however it’s pressurised because it comes either from a tank or mains. We know because of our earlier discussion that this means it will have a higher boiling point than the water in the tank. This means that you can never boil the water in the copper pipe. So we now have a system that can’t go boom.

Now a couple of tips. You will have to periodically (weekly or so) open the drum fill tap to top up the drum as expansion and evaporation will continually lower the level in the drum. You could set up an external float system to keep the drum topped up but it’s more complexity and I haven’t bothered. How often you fire the system depends entirely on your usage. The PRI system was fired once in the morning to top up heat lost during the night and to cover the small number of people that preferred morning showers and again in the evening, beginning an hour or so before showers with people throwing the odd stick in as they came. The initial fire up of the system brought 180 litres of water to boiling in just over an hour and consumed approximately 4 kilos of scrap timber.

A few words of caution — there are home built wood powered systems around using complete old hot water tanks, but I cannot recommend using these if the system depends on the original safety valve. A wood fire is very hard to precisely regulate so invariably at some stage the water in the tank will boil. Electric or gas powered systems are designed with thermostats that precisely regulate the water temperature. The safety valves on these are there purely for if the thermostat has failed, which would allow the water to boil. The safety valve works two ways, for excessive pressure and for excessive temperature. It is designed to fail (yes fail) if either of these parameters are breached. Once this happens it will leak. In a pressurised system this means it will dribble out water non stop till some smarty pants comes along and blocks it, which then converts it into a bomb. So, if you have one of these systems you’re literally dicing with death.

Contrast this with the PRI system that has no valves to fail and has been designed so that it pretty much would require a failure of the laws of physics for the system to get to an explosive state. If you require even more safety — and, frankly, I haven’t bothered because the laws of physics seem quite stable at the moment — you can arrange things so the overflow outlet is directed towards the rocket stove inlet (not directly above of course) so that if and when the water boils steam blows into the rocket stove, putting the fire out.

On my next post I’ll be talking about how we can reduce the energy needed to heat water by as much as 30% . Hint, what happens to the hot water after we spend so much time and effort getting it up to temperature?


Tim Barker is teaching workshops at Koanga. Learn more about rocket stoves, natural building and appropriate technology and get some practical experience at our Autumn Internship and our Appropriate Technology Workshop. Book now for early bird discount!