Updated in February, 2012 from an article originally published in Ceramics Monthly (www.ceramics
monthly.org), September, 1990, Pages 44-50. Reproduced with permission. Copyright, the American Ceramic Society.

Author Bio

the author


The design, construction, and firing of any gas kiln can be very hazardous. The concepts described in this article should not be attempted unless you are experienced and knowledgeable about the risks of gas fumes, fire hazards, and explosion hazards related to firing with gas. What is described here is experimental. As such, there may be unexpected occurrences.

My kiln is outside so that gases do not accumulate. It has a metal roof to protect it from the weather.

authored by
Marvin Bartel

Bowl Assignment

Castable for Soda Kiln
Abstract Expressionism Clay Lesson

Learning to Throw Pots: Tutorial
Ceramics Courses

Ceramic Links

Clay Lessons for Schools

Bartel Clay Recipe

Coil Building Assignment

Cleaning & Working Safely

Cylinder Assignment

Hazards in Ceramics

Hints for Potters

Clay and Kids
Clay for Toddlers

Photo Emulsion you make

Process your own clay
Rework old clay

Reprocessing Old Clay
Demo or Hands-On Learning?
Learning from the Clay - video
Marvin Bartel Home Page

Marvin Bartel Pottery

Bartel Exhibition Catalog

Ceramics Alumnus, Tom Unzicker
Tom & Jeff Unzicker Pottery

Goshen Alumnus, Dick Lehman

Contact Marvin Bartel

Goshen College Home
Art Department Home
Art Education by Bartel


How different is this kiln?

I built the first fuel efficient kiln in 1975. It continues to amaze me that so little heat escapes from the chimney. Early during the first firing, to clean up the site, I tossed a paper cup into the chimney, assuming it would burn. After the kiln had reached cone 04 for a bisque firing, it was turned off. A plume of black smoke suddenly rose from the chimney. Scurrying up a ladder, I saw the cup just beginning to collapse in flame. It had lodged inside the chimney, but hadn't gotten hot enough to burn until the system was turned off. It was obvious we had a kiln unlike any I'd ever seen or heard of. Today visiting potters can't believe it when they see hickory and oak leaves on the trees so near my kiln chimney.

How is this kiln different?

It used conservation of heat. According to Francis(1) 74 percent of the heat during a typical cone 10 firing escapes up the chimney. This leaves a mere 26 percent of the fuel's energy to heat the ware and provide all the heat loss through the kiln walls. What potter hasn't experienced actual temperature decline while firing? Some of this energy loss is the result of an improper gas-air ratio, inadequate fuel supply, the thermal mass of kiln materials, and/or poor kiln insulation. Still, much of of the firing cost is expended in the heating of cool air coming into the burners. It requires most of the energy of the fuel just to heat the incoming combustion air. When the kiln is very hot, it is hard to raise the temperature to a higher temperature when the combustion air is entering the burners at 70 degrees F. 

In industry, some applications use pure oxygen to avoid some of the cooling effect of using air.  Gas welders use pure oxygen. In kilns, we allow air to supply all the needed oxygen from free air, but in this design, I conserve the heat that would otherwise be lost with the flue gases.

Firing speed is another important fuel use variable. Faster temperature increases can cost less because all of the ways that heat is lost are related to duration. Longer firings require more fuel because their is more time for heat loss to take place. Safe firing speeds depend on the pottery in the kiln and the kind of kiln materials being used, but that is not the subject of this essay.

Even with perfect combustion and no heat loss to the walls, 74 percent of the heat escapes through the chimney regardless of how much fiber and softbrick is used. Consequently, the best insulated fiber kilns are still heated with only 26 percent of the fuel's energy.

How does it work?

I force normal air through many refractory pipes that are placed in the chimney. The chimney is a flat slit all across the back of the kiln. The flue gases with their waste heat travel up the chimney (around these rows of refractory pipes). The chimney incorporates rows of cross-flow or U-shaped air tubes carrying combustion air downward in successive steps. Near the base of the chimney the air is channeled to a combustion chamber. At this point the gas also enters the combustion chamber. Hence, there is no cold combustion air coming into the kiln or burners. Much less heat leaves at the top because the refractory tubes absorb the heat to heat the combustion air. Very little of the kiln heat leaves the chimney. Very little fuel is needed to increase the temperature of preheated combustion air.

Where did this idea originate?

I was inspired by traditional hill-climbing kilns in the Orient. They are efficient because latent heat form downhill chambers (even after these chambers are finished firing) preheats combustion air for uphill chambers. The top chamber in six chamber hill-climbing kiln fires from cone 04 to cone 10 with a small fraction of the fuel needed to fire the first chamber from cone 04 to cone 10. I set about designing a self-contained single chamber system because I knew that I and most potters cannot build or use a six chamber hill climbing kiln.

How many of these kilns have failed?

I have built four kilns using this system. The first kiln had some problems that were solved by providing larger combustion space. The combustion area was originally under the kiln floor. It was redirected behind bagwalls along each side. This solved the problem and the kiln is still in use.

The second kiln failed for two reasons. Originally, we used inadequate refractory for the heat exchanger tubes. Secondly, we failed to allow enough combustion space and used inadequate refractories for the higher flame temperatures produced by the preheated combustion air. The kiln owner abandoned the concept and used conventional burners.

The first kiln, my studio kiln, built in 1975, uses 50 tubes (two inches in diameter) in the chimney and is still in regular use. The heat exchanger tubes were originally extruded as mullite muffle kiln flame tubes for kilns built by Denver Fire Clay Company. The third kiln, at Goshen College, fired nearly all the student work from 1977 to 2007. It used 150 mullite tubes in its chimney. These are one inch slip cast tubes purchased from a refractory company in Illinois. The fourth kiln I built in 1985 has twenty five mullite two-inch U-shaped tubes in the chimney. We made these tubes with an extruder and pre-fired them to cone 14. The kiln also has a secondary heat exchanger fabricated from sheet metal above the chimney to extract residual heat. The combustion air is first directed through this steel heat exchanger. Then it goes through the primary ceramic heat exchanger in hotter portion of the chimney. More details follow.

What is the unique shape of the chimney?

The chimney in these kilns is a four inch wide slot all across the back of the kiln. This shape provides the necessary support for the tubing and adequate space for the flue gases to flow freely over the tubing.

A heat exchanger harvests the waste heat inside the kiln chimney. These refractory tubes in the chimney heat the combustion air. Above, the straight tubes are being mortared into my kiln in Goshen, Indiana. These tubes were installed in 1975 and are all still in use in my kiln in 2012.

Counter flow vs. parallel flow hear exchangers?

I have used only counter flow heat exchangers. In a counter flow design the hot gases from the kiln move in the opposite direction from the incoming combustion air. The counter flow design is the only configuration offering the possibility of nearly 100 percent heat recuperation. Because of the higher working temperatures, a counter flow heat exchanger must be made of high temperature materials with good resistance to thermo shock.

In a parallel flow design the cold air is brought into the heat exchanger at the hottest point. The two temperatures average out. Hence, the preheating potential of a parallel flow design can never retrieve more than half of the heat being wasted via the flue. I see no point in a 50 percent potential, when a 100 percent potential is possible, less heat resistant materials can be used in a parallel flow design.

Our first two kilns used purchased straight mullite round tubes. One of these kilns is still firing with the original tubes. Assembling straight round tubes requires very careful and complex cutting and fitting of soft brick and lots of kiln mortar to fabricate the required manifolding. Furthermore, in the event of tube breakage, a major tear-down and rebuilding is required.

Therefore, in subsequent kilns we've used U-shaped square tubes in the chimney. This allows easier construction and access to inspect and replace parts in the event of breakage.

I built a kiln like this at James Madison University, Harrisonburg, Virginia, in 1985. Its usable firing chamber is 27 cubic feet. Fully loaded, the first firing used $10 worth of gas to fire it to cone 11 at a price of 50 cents per CCF (hundred cubic feet) of gas. It was a slow firing to dry out the kiln parts.

U-shaped tubes are extruded using an offset die on an expansion box. The offset produces the desired curl. Extruders can be built or purchased from Bailey Pottery Equipment, Kingston, New York.

A durable material for heat exchanger tubing has been hard to find. I began by using a flameware body, thinking its thermo-shock characteristics would be essential. However, expecting flameware clay would melt, I added alumina. It still failed after a few firings.

When the kiln is at 2,300 F, the combustion air comes in at 2,000 degrees F. Above, the U-shaped tubes are being mortared into the kiln at James Madison University, Virginia, in 1985. I asked about it about 15 years later, and it was still working. So far as I could learn, it is was being used anymore in 2009.

We now make the U-shaped tubing from a mixture of one part OM4 ball clay and three parts kyanite from the Virginia Kyanite Corp., Dillwyn, Virginia. The ball clay adds enough plasticity to make the mixture extrudable and provide green strength. The high alumina content of the kyanite forms mullite when fired. Mullite is resistant to both high temperatures and thermal shock.

The tubes are fired with stoneware to cone 9 or 10. Additionally, I fast fire the tubes to cone 14 in about 2 hours. I use a small kiln formed in an oil drum using four inches of fiber. I'm not sure cone 14 firing is essential, I started the cone 14 firings when I was trying to make the flameware durable enough, but this mixture is not flameware. It is a high mullite clay.

While silicone carbide may be a more reliable material and a better heat conductor, I have not found a company willing to manufacture U-shaped tubes. The cost of silicone carbide tubes would be higher, but it may be worthwhile in the hottest area.

For the secondary heat exchanger above the chimney I use stainless steel sheet metal in relatively thin 26 gauge to minimize cost and facilitate good heat transfer. Stainless steel holds up because the flue gases have been cooled by the primary heat exchanger. Drawing no. 6, below shows details of the stainless steel heat exchanger used on top of the kiln chimney. It is the one in use at James Madison University.

Below, the metal heat exchanger serves as the top portion of the chimney. It harvests heat that the refractory heat exchanger below misses. Here it has paper cups on top that have turned black during a cone 11 firing, but they did not burn until the system was turned off, proving that most of the heat is being cycled back into the system via the combustion air.

This photo was taken at the end of the first firing at James Madison University, 1985. The three blowers are on the back side of this metal box. It has seven vertical stainless tubes. The exterior of the box is galvanized sheet metal. Inside the shell there is an inch of fiber insulation. There are two stainless air baffles that move their air back and forth across the seven vertical flue tubes.

What do the burners look like?

Forget burners. This is "outside the box" thinking for potters. Gas won't premix with 2000 degree F air. It ignites on contact with the air this hot. Like a diesel engine with fuel injection and a turbocharger, there is no premixing of fuel and air. The raw gas and the hot air come directly into a common combustion chamber. Experimentation with the bag wall gives uniform firing temperature and protects the ware from raw gas. To protect clay and glazes from strange effects from raw fuel, there should be no openings in the bag wall in the vicinity where the gas and hot air first meet.

The tip of the gas pipe is extruded from the same material used for the heat exchanger tubing. The mullite tube is telescoped over a stainless steel gas pipe inside the kiln wall with ample kiln mortar. The stainless steel pipe is pre-connected to flexible gas tubing to isolate it from any movement that might loosen its bond with the kiln. The stainless flex tubing sold in hardware stores for connecting gas kitchen ranges provides a flexible connection.

Note: Stainless steel flex tubing was not used in the above photo, but a loupe of it would fit at the point called "main gas input" in this photo.

A gas-tight seal of silicone caulk around the pipe is used on the kiln exterior to further prevent leakage. There must be no orifice or other restriction inside the kiln to produce back pressure on the joint between the pipe and the kiln. Everything is checked for gas leaks with soap suds.

Not knowing the orifice size needed, we include a final gas valve in the piping near the kiln. During the final stages of the first firing it is set as small as possible, while still achieving an acceptable temperature rise and reduction atmosphere when desired. Once the setting is determined, the valve is ignored. It essentially acts as an orifice would (producing measurable pressure in the gas line) on a conventional burner.

All subsequent adjustments are made with the valve upstream from the pressure gauge. This pressure gauge gives a reading to record in the firing log, helping find and record the most effective gas-air mixtures.

How do you monitor the flame for safety?

I monitor the flame with an ultra-violet (UV) flame sensor. Potters affectionately call them purple peepers. A UV sensor is located on the cool end of a pipe aimed at the flame. If it doesn't see flame, it shuts everything down. See drawing and photo above.

You'll note the drawing includes a pilot. The pilot "candles" the kiln overnight to dry the ware before turning on the blowers and the main gas supply. Once the kiln is firing on the main flame, the pilot is turned off and its air port closed. The same UV sensor monitors the pilot and/or the main flame.

What about blowers?

Since the flue is cooled, it won't draw on its own. Forced air is required. Shaded pole motors with squirrel cage fans are quiet and cheap. For a forty cubic foot kiln I use a Dayton, Model 3M601, blower equipped with a 5.8 amp 115 volt motor operating at 1050 RPM. It uses less than a dollar's worth of electricity per firing.

On a 25 cubic foot kiln we used three smaller blowers (Dayton model 4C013A) with anti-backflow flaps added. Even if one blower fails, the firing continues, albeit reduced.

Some shaded pole blowers respond poorly to varying their speed. Therefore, I provide a way to dump air. An adjustable hole near the top of the heat exchanger allows air to escape during early stages of firing. I prefer to run blowers at their intended speeds with an unrestricted intake to minimize stress and overheating.

On my kiln I mounted the bower is on hinges. I tip it up when less air is needed, dumping some air outside the system. The blower hangs by a chain tipping the whole unit up. Most of the air misses the heat exchanger during the early phases of firing. As the firing progresses the blower is tipped closer each time the gas is turned up. I record chain link settings, gas settings, pyrometer readings, and the time in a firing log.

It is possible to calculate the rate of temperature increase within minutes after an adjustment. I soon learn the most effective settings at each kiln temperature.

Finally, seal the kiln well. If you blow the heat out around the kiln door, the benefits of the heat exchanger is obviously lost.

What about overfiring the kiln?

With my old hardbrick fuel-guzzling kiln my children and I went fishing at Emma Lake one day. The fish started to bite and I totally forgot the kiln was due to be turned off. It fired an extra four hours. I got a few runny glazes, but the kiln wasn't damaged. The crappie and bluegills weren't bad either.

A good heat exchanger in the chimney increases the effectiveness of the fuel. Twenty extra minutes significantly overfires the kiln and four extra hours could seriously damage the kiln. I'm not just talking a few runny glazes, but melted bricks!

As "absent minded potter" protection, I install a Dawson Kiln Sitter near the bottom of the kiln. To avoid early shut-down nuisance, I trigger the kiln sitter with a cone rated slightly hotter than I plan to fire. This provides cheap kiln insurance for absent minded potters. It has to be near the bottom of the kiln to avoid the reduction firing soot that can jam up the mechanism.  A modern computerized kiln control may be a better option.

What about other fuels?

We have used some wood in one of our kilns. The preheated combustion air is wonderful. Saw dust or wood chips would be ideal for potters in an area with a free source. I have added pellets made from sawdust. I tried adding oil, but abandoned it because of too much black smoke. Theoretically, any combustible material benefits from this system. 

I feel that electric firing is among the least efficient and most polluting unless the power is produced from wind, solar, hydro, etc. In Indiana, we get nearly all of our electric power from coal. Generally, less than 35 percent of the power used at the generating plant ends up at the kiln. Production and transmission inefficiencies waste 65 percent or more. Acid rain, the greenhouse effect, and global warming result. Particulate matter and mercury from coal plants also contributes to health costs and shortens lives.

I'm hoping my experiences will give you some ideas and perhaps inspired some idealism. We who work in the pottery village of today's earth need to become more creative and whole. We need to work together to create the new "hill-climbing" kilns to close our holes of waste.

These are the folks who built the kiln at James Madison University

What are the next steps?

I have not been able to test hydrogen, but I believe it could be used in place of gas. I intend for my next kiln to use more fiber so it has less thermal mass. It will include airspace in the exterior skin of the kiln so that the combustion air begins to be warmed by heat that passes through the kiln walls. I believe that the outside wall of a kiln should be cool enough to place your hand on it when it is cone 10 inside the kiln.

I'm hoping my experiences will give you some ideas and perhaps inspired some idealism. We who work in the pottery village of today's earth need to become more creative and whole. We need to work together to create the new "hill-climbing" kilns to close our chimneys of waste.



(1). Francis, Buddy. "Heat Recovery" published in chapter six in a book by Brodie, Regis C. The Energy Efficient Potter, Watson-Guptill, 1982. p. 159.

(2). Moore, James W. The Changing Environment. Springer-Verlag, New York, 1986. p. 206.

Other References:

Colson, Frank A. Building Kilns With Space Age Materials. Van Norstad Reinhold, New York, 1975.

Olsen, Frederick. The Kiln Book. Keramos, Bassett, CA, 1973.

Rhodes, Daniel. Kilns: Design, Construction and Operation. 2nd ed. Chilton, Radnor, PA, 1981.


ALL RIGHTS RESERVED: These photos, drawings and text are copyright.  Reproduction and publication other than scholarly review is not permitted without permission from the author and from Ceramics Monthly.  You may link your page to this page.

April 20, 2009 update