Biofuels | Wind | Solar | Hydro 

Home | Search | Subscribe

Wood Heat | Sun on Tap | Wood Fired Hot Tub

A Better Barrel Stove | Fire and Water Part 1

The Cost of Heat

Fire and Water Part 2

Building A Barrel Stove Hot Water Heater

By Robert Saunders

There is a real creosote problem in the heat exchanger and the stack.  After less than a month the buildup on the coil and in the stack is unacceptable, both from a risk of fire and from a problem with heat transfer in the coil.  I have a tentative solution for both problems, but I will have to build and test a new heat exchanger and stack. 

Test Report #1

When I read Fire and Water, by Art Sussman and Richard Frazier, on this web site (, I decided I had to build one of these things. It helped that I already had a barrel stove. The whole project started with an attempt to cut down a very large monthly electric bill, especially in the winter. This approach appealed to me because it was lo-tech, simple and relatively inexpensive.

I used the information contained in the article as a guide and built the barrel stove hot water heater described below. I am pleased to say that it works as advertised. So far the cost for the system, as illustrated in Figure 1, is less than $200.00. I expect to save at least $50.00 per month by operating it every 2nd or 3rd day for hot water in the summer. In the winter it will run continuously when I expect to save more than $200.00 per month for heat and hot water. Not counting my labor and research, thatís a pretty good investment.

Figure 1. Basic Barrel Stove Hot Water System

The Basic System.

Three basic units make up the heating system illustrated in Figure 1. It consists of the Barrel Stove, the Heat Exchanger and the Hot Water Storage Tank. The barrel stove and the storage tank are readily understood and are explained in some detail in Fire and Water. However, details of the heat exchanger are not as well described or documented. A brief search for a heat exchanger of the type described in the article failed to locate any commercially available units. Therefore, I was forced to design and build my own heat exchanger. After buying some materials and special tools, this DIY effort resulted in a successful design that yielded the following system performance.

The inlet water to the bottom of the heat exchanger is cold water (condensing atmospheric humidity) from a well. I estimate the inlet temperature is about 40 degrees F and the outlet temperature from the top of the coil is about 130-150 degrees F. I intentionally kept the fire low to avoid building up any steam in the system, although there is a safety relief valve in the tank side of the tube from the stove to the tank. The relief valve is pre-set to open at 210 degrees F or 150 PSI. The whole system was built outdoors for testing. Only the pipe from the top of the coil to the tank needed to be insulated.

The Barrel Stove.

The barrel stove is really interesting. It's one of the most common and oldest wood burning stoves around (and least expensive). They are and have been very common in Alaska and northern Canada. They're a little tricky to operate but you can burn anything in them. They can operate at high temperatures where the barrel gets red hot. However, the kit manufacturer recommends against this kind of overheating. One problem is radiant heat when they get that hot. Another problem is smoke. They can be very smoky since there are no baffles in them, resulting in a lot of unburned gases. Aside from the disadvantages, they can generate lots of heat. When damped down, they can burn for hours without adding any additional wood.

System Components.

The basic system consists of the following components:

* Wood burning barrel stove
* Hot water tank
* Stack or chimney
* Heat exchanger
* Interconnecting pipes

Theory of Operation.

Before you think about building a similar hot water heating system, you should read the referenced article Fire and Water by Sussman and Frazier. They have covered the basic theory of operation of the system in broad detail. I will limit the discussion in this article to the application using a barrel stove for heating hot water. At some future date I will expand the application to include heating with hot air or hot water from the same system.

The Barrel Stove Kit.

If you are planning to build a barrel stove, there is one standard inexpensive barrel stove kit available from many local hardware stores. It is manufactured by Vogelzang Corporation. They provide detailed instructions on building and operating various models of barrel stove on the following web site.

This web site is especially important for its detailed list of requirements necessary to safely operate the barrel stove. When buying a kit, the cost of buying one locally or by catalog may be the same, after shipping charges are added in.

A second source for Barrel Stove kits can be found at the following web site:

The Barrel Stove consists of a barrel stove kit, a steel barrel and a set of stovepipe sections, which I call a stack. I used a 55-gallon steel drum (a tight head steel barrel, i.e., not the type with a removable top and separate locking ring). The barrel diameter is approximately 23 inches.

The kit consists of a door assembly, front and rear leg assemblies and a collar and damper assembly. The door assembly I used includes a sliding vent. Some models include a dial type vent. Although the dial vent is more convenient, I prefer the sliding vent because it allows you to see the glow from the fire inside without opening the door.


A double barrel kit is available for use with two barrels stacked one above the other. The main advantage with this design is that less heat is wasted in the stack and more hot surface area is available to heat the space being heated. However, there is probably no advantage in our application.

The following web site offers experience with another variation of a barrel stove design.

Building the Stove.

I purchased a stove kit for about $45.00 consisting of a door assembly with a sliding vent, front and rear leg assemblies, and a collar & damper assembly. I was referred to a local welding shop who could provide me with a barrel and who agreed to assemble the whole unit for $55.00. The only other parts I needed were three 24Ē sections of 6Ē stovepipe, one 24Ē section of 8Ē stovepipe, Two 6Ē to 8Ē reducers, a 6Ē adapter and the 6Ē rain cap. I also needed copper tubing and fittings.

The Heat Exchanger.

In this first model, I installed a coil of 3/8" copper tubing in the first stovepipe section in the stack. I used an 8Ē section of stovepipe and two reducers on either end for the heat exchanger. This allowed me to make a larger diameter coil without kinking the tubing. It also provides more tubing surface area exposed to the hot gasses. If you are interested in more details on the construction and performance of the heat exchanger design, you can contact me at the following address :

The heat exchanger I built works like a charm. There is a problem with putting a heat exchanger in the stack, which has to do with cooling the gas temperature in the stack and the problem of creosote. In practice the creosote is supposed to run down the stack and burn in the firebox. The problem comes when it cools off and sticks to the surface of the stack. Sooner or later it builds up and blocks the chimney or stack and will cause a fire if it isn't cleaned out. It's somewhat evident in my stack, the top of which is only about 12 feet high. The higher the stack, the more of a problem it becomes.

There is a real creosote problem in the heat exchanger and the stack.  After less than a month the buildup on the coil and in the stack is unacceptable, both from a risk of fire and from a problem with heat transfer in the coil.  I have a tentative solution for both problems, but I will have to build and test a new heat exchanger and stack. 

Assembling the Stack.

I had some problems assembling the stack. The standard procedure is to place the crimped end of a stovepipe section down into the smooth end of the lower stovepipe section. However, the stove kit collar is machined on the outside to mate with the smooth end of the first section. The 6Ē adapter fits the collar but it allows creosote to leak slightly. This can be sealed with stovepipe cement, but I didnít bother. Eventually it sealed itself with solidified creosote. The other sections fit properly. I used three self-drilling screws, spaced about 120 degrees apart, to secure each section. The screws securing the adapter to the cast iron collar worked but they didnít screw in all the way. To prevent cracking the cast iron collar, I didnít try to force them any further. I added the rain cap to prevent rain from entering the stove. The cap also prevents wind down drafts from entering the stack and choking the flow of gasses from the stove.

This completed the assembly of the stove. I tested the stove by burning kindling , wood scraps and logs in it and later added the heat exchanger, after I read Fire and Water.

Operating the Barrel Stove.

The door opening to the barrel stove is approximately 10 inches square and will easily take 8-inch diameter logs up to 24 inches long. After I start a fire with paper and cardboard and kindling, with the vent and the damper fully opened, and it has built up a pretty good layer of hot coals, I add three logs, two on the bottom and one stacked in the middle, like a pyramid. Then I close the vent and damper somewhat. At this point the damper is at about 45 degrees and the vent is open enough to see a glow through the openings. I try to check the fire every hour, observe the smoke, and adjust the vent until Iím satisfied that I have the fire the way I want it. Depending on the size of the logs and the amount of air getting in through the vent, the fire should continue to burn overnight. The final adjustments of the vent and the damper, and the amount of smoke, are going to depend on how hot a fire you want, on the type of logs being burned and how well they are seasoned. In general, hard woods, like oak, burn hotter and produce less smoke. Soft woods like pine and birch burn slower but produce more smoke and creosote. The type of wood effects the amount of venting and damping needed.

Building a Test Stand.

I designed the test stand to be certain it could hold a 500-lb. tank filled with water. The frame is constructed with 2x4s and the sides are cut from ĹĒ OSB sheeting. Each corner has two 2x4s with two studs in the middle. The top is made up of 2x6s cut to length with a 1Ē overhang on each side. I left a nailís width spacing between boards on the top. I made it 3í wide and 6í long to give me a platform to stand on while I worked on the tank. The sheeting gives the whole structure stiffness so it wonít collapse. Itís all nailed together with 10d nails.

Locating the Test Stand.

Nothing flammable should be located within three feet of the barrel stove. I placed the stand three feet from the corner of the stove as shown in the following drawing. This shows the areas of highest intensity radiant heat from the hot surfaces of the stove.

Figure 2. Test Stand Arrangement


Finding a Hot Water Tank.

I took the advice given in Fire and Water and called my local plumbing supply place. They had a number of tanks in their yard waiting to be picked up for disposal. I picked the one that looked the newest and that had side openings on the top and the bottom. Although it looked new, it was actually more than ten years old. It is well insulated. (I can barely feel any temperature gradient along the surface of the tank when itís filling with hot water.) The only thing I found wrong with it was a blocked hot water check valve. However, I haven't hooked up the power to it. That would seem to defeat the whole purpose. It's rated at 4500 watts.

Preparing the Tank.

Before I could connect the tank to the stove, there were a few things that needed to be done.

Domestic hot water heaters are connected to the plumbing in the house by the use of unions in the hot and cold water pipes at the top of the tank. When I got this tank, the lower half of the unions were still attached. The unions were attached to the hot and cold water check valves. These valves look like 2Ē nipples but are actually ball check valve assemblies. They are designed to prevent the flow of water in the opposite direction from the normal flow of water. Both check valves are identical and must be inserted in the right direction. One is inserted inverted from the other. If either check valve is inserted backwards, water will not flow into or out of the tank. Itís a good idea to inspect the check valves for proper operation. If you shake the valve the ball should be free to move. In my case, the hot water check valve was blocked. After I flushed it out, it worked perfectly. In some cases a used tank might be blocked with rust. The best option in that case is probably to look for another tank. Unfortunately, there isnít any easy way to inspect the inside of the tank. If the water flows when you fill it, it is probably good enough for this application. It is likely that the tank was discarded because of a failure in the electrical system, either the heating elements or the thermostat control, or in the gas burner.

Fortunately, this tank had a relief valve inserted into the top opening in the side of the tank. I removed this safety valve and inserted a tee and nipple in the opening. I reinstalled the valve in the tee and added an adapter to the remaining opening in the tee to connect to the copper tubing from the top of the heat exchanger coil.


Whenever any connections are broken, it is necessary to clean the threads and apply Teflon tape before mating the joints.

Connecting the Tank to the Heat Exchanger.

I used two 6-foot lengths of 3/8Ē copper tubing, with compression fittings on either end, to connect the upper and lower fittings of the heat exchanger to the upper and lower connections in the side of the tank. I bent the lower tubing so that it has wide bends and lies horizontally between the lower connections. The tubing is stiff enough not to require any supports. The upper tubing is about the right length not to have any sharp bends in it. I added ĹĒ foam pipe insulation to the upper tubing.

Flow Rate.

I don't have a pump in this setup. It relies on thermosiphoning, which is a fancy term for the concept that hot water rises in a pipe and, when properly connected to the top of a tank, is replaced by cold water from the bottom of the tank. I'll get a better idea of the flow rate later. I can time how long it takes to fill a gallon jug and plot some temperature curves.


Some temperature measurements can be estimated without an instrument. Of course the simplest instrument used for measuring temperature is the thermometer.

* Water splashed on a surface will freeze if the surface temperature is below 32 degrees F (0 degrees C)
* Water splashed on a surface will boil if the surface temperature is above 212 degrees F (100 degrees C)
* Water condenses on a surface if the temperature of the surface is below the dew point. The dew point can be found from a local weather report.
* If a surface temperature is just hot enough to be uncomfortable to hold a hand on it for more than a few seconds, the temperature is near 130 degrees F, or is said to have a 60-degree rise. Motors are sometimes specified by its operating temperature in terms of a temperature rise above an ambient temperature of 70 degrees F.

I use the last method for estimating the temperature of the hot water without an instrument. Iím confident that if the temperature of the copper tube at the top of the coil is too hot to hold my finger on it for more than a few seconds, the fire is just right. This should be between 130 and 150 degrees F. I always wet my finger before I touch the tube. If it sizzles, itís too hot. If the temperature of the water in the tube exceeds the boiling point and turns to steam, the relief valve opens. In this case, I damp down the fire by reducing the flow of air through the vent to reduce the fire and lower the output temperature. Remember that the amount of fire in the stove is controlled by the amount of air available for burning and not by the amount of wood in the stove.

I plan to get better estimates of the temperature and flow rate of the water from the heat exchanger. There are some inexpensive ways to obtain accurate temperature measurements. An informative discussion of available inexpensive temperature sensors may be found in the following article and in comments from its readers:

DIY Temperature Sensors
by Steve Spence

The Next Generation.

Eventually, I plan to house the whole system in a steel shed (cost about $500) and run hot air ducts or hot water pipes to heat the house, but thatís down the line. I plan to work on a dual coil heat exchanger. The additional coil will provide hot water to heat at least one room in the house in the winter. This setup will require a circulation pump and a radiator in the room being heated.

Copyright © 2003 by Robert Saunders



Click Here!