Downdraft Kiln Conversion Boris Robinson

Appears in the April 2020 issue of Ceramics Monthly.

My initial exposure to making pottery was in a well-equipped museum studio that had two large gas kilns. When the venue faltered, my only option was to go to a commercial studio that fired to cone 6 in electric kilns. I tried it, but felt something was missing. I had begun a path that led to studying the influences of Korean, Japanese, and Chinese pottery and I really wanted to continue with this exploration of high-fire clays, simple glazes, and reduction-firing techniques.

The easy solution, so I thought, was to purchase or build a downdraft gas kiln. Searching to buy a kiln like the smaller one at the museum, it quickly became evident that rectangular gas kilns were above my budget. Another constraint was that I wasn’t sure where I could install the kiln and for how long, so it needed the ability to be moved, thus making large, external firebrick chimneys or structures impractical. This also meant I wasn’t going to try and build a large rectangular gas kiln as detailed in several kiln-building books. Having eliminated these other options, my primary focus became converting a used electric kiln to gas downdraft.

One of my museum class teachers, Walford Campbell, was using an old eight-burner updraft kiln and was also interested in a downdraft kiln. Walford provided a wealth of knowledge in helping me to understand the fundamental aspects that would make a small downdraft kiln function properly. The result was a downdraft kiln design that works incredibly well, has a very basic construction, and is simple and inexpensive to build. The kiln design provides adequate interior space, even temperatures, repeatable firings, and solid reduction. The prototype, built in 2017, was a small 23×18-inch-high electric kiln, the second a 23×27-inch electric kiln, and the third was Walford’s 28×32-inch updraft kiln converted to downdraft.

Draft

Marc Ward of Ward Burner Systems (wardburner.com) wrote that, “Draft is the life’s breath of a gas-fired kiln.” This statement is so simple and seemingly obvious that the idea gets lost amid the physical aspects of the kiln construction in most electric to gas conversions; however, it is critically important. Draft is the flow of air and burning gases through the kiln. In a chimney/flue, the velocity of the airflow is a function of the difference in temperature between the hot gas and air inside the kiln chamber and the outside air. The hotter the air in the kiln is, the lower its density and pressure will be relative to the cooler outside air, thus, the faster it will rise. The movement of the hot air going up the chimney causes a suction that quite literally pulls the gases through the kiln.

Specifically, when air is heated it rises past surrounding cooler air. For instance, when an oven door is opened while baking, the hot air in the oven rises quickly upwards. Hot air has less density than cooler air and this essentially means that hotter air is lighter than cooler air. Less dense, hotter air in a kiln wants to rise in the same manner, and when the hot, rising air is confined within a chimney, it essentially acts like a piston moving upwards and pulls the air below it up. This is called the chimney effect, and it is what creates the draft and flow of the air through the kiln. A taller chimney allows the hot air more time to rise and accelerate to a faster speed and this causes more air to be sucked through the kiln.

Several books on kiln construction suggest a starting point for chimney height of approximately three times the inside kiln height to adequately regulate airflow and draw. Chimneys cool the gases, resulting in a lower pressure difference and a lower upward velocity, and thus chimneys need to be tall enough to create the adequate suction.

Electric to gas conversions I observed usually had one consistent build method, which was to use firebrick for the interior flue construction. Besides taking up a lot of space, the bricks insulate the gas in the flue from high temperatures in the kiln, contributing to the gas cooling and its velocity slowing. In most cases, the short height of a typical electric kiln with an interior firebrick flue isn’t tall enough to provide enough draft suction to reach the higher cone-10 temperatures. A common issue is kiln stalling. As the kiln interior gets hotter during firing, the combusted propane/air mixture coming into the kiln has a greater expansion, increasing the gases’ volume. If there is a larger volume of air coming into the kiln chamber than the volume able to go up the flue and out of the kiln, then the kiln has reached its maximum temperature for the hot air to properly flow through and burn, and it will stall. A further increase in temperature causes more gas expansion and requires more air and gas flow up the flue than the flue can handle. Attempted fixes to stalling usually involved adding a chimney or metal flue pipe on top of the kiln lid to create more draft suction, therefore relieving the constriction.

When converting an electric kiln to a downdraft gas kiln, a simple and effective design solution to counter the short flue heights is to provide a heat source to the flue gases in order to create and maintain a higher draft velocity. A thermal draft inducer was made by using a thin cordierite kiln shelf to create a flue wall inside the kiln, as shown in image 5. Draft inducers are used in industrial and home-heating applications and a thermal draft inducer transfers heat to the rising flue gases in order to maintain their lower density and higher velocity.

One cubic foot of propane produces around 2500 BTUs and requires approximately 25 cubic feet of air for proper combustion. Near cone 10, both burners in the 28×32-inch kiln are producing about 150,000 BTUs using 60 cubic feet of propane and 1500 cubic feet of air, for a total of 1560 cubic feet entering the kiln per hour, or 26 cubic feet per minute. Since the heated gas expands over 5 times its initial volume near cone 10, the required flue output is about 140 cubic feet per minute. This is a lot of gas flow and inadequate flue size can contribute to the reason many converted kilns stall. The 28×32-inch kiln’s flue is 2 7/8 inches wide with a cross-section area of 35 square inches and near cone 10, the flue velocity equates to about 10 feet per second (7 mph). With the thermal draft inducer, the flue velocity is sufficient to create enough draft suction that additional chimney height is not needed. For a kiln with different measurements, use online calculators to determine adequate flue area and dimensions. You will input the air density, air temperature, differential duct areas, and duct length.¹ Note that other factors like shape and burner-port placement could affect draft, so testing is necessary. 

For the 28×32-inch kiln, 5/8×20×20-inch cordierite shelves were used for the flue wall, and ¾×16×16-inch shelves were used for the 23×27-inch kiln. The lower flue-wall shelf was placed on a 2-inch kiln post to create the bottom flue opening. The flue walls and ware shelves were cut using a cheap wet-tile saw. Vertical notches were cut in the soft firebrick kiln wall into which the shelves were pinned with short bits of twisted electric-element wire pressed into the bricks. A mix of refractory cement and ceramic fiber was then pressed into the spaces and over the wire to seal the shelves in place. I used approximately equal amounts of Laguna Smoothset Mortar and ceramic fiber cut into small bits and added water to a make toothpaste-like consistency. On the other side of the kiln, a 10-inch tall piece of thin kiln shelf is used to separate the burner flame from the gases descending around the bottom shelf to the flue. This burner wall was not pinned into place, and so a 1×3½×5-inch tall piece of soft brick is used to ensure that the burner wall does not lean into the burners.

Design Specifics

There are a few minimum design considerations specific to this conversion method that if changed may impact the firing performance. First, affordable MR-750 Venturi burners were used rather than cheaper weed-burners or the more expensive forced-air burner systems. The brass orifice that comes with the MR-750 is too large, but new orifices (drilled with a #50 drill bit (0.07 inch)) can be ordered from Ward Burner Systems. The smaller orifice size increases the velocity of the propane gas in the burner and also allows for a higher pressure and finer flame adjustability. The burner is placed below the kiln pointing upward. The burner hole in the kiln floor is 3½ inches in diameter with the top of the MR-750 placed ½ inch to ¾ inch below the floor. This size and placement provide for the proper ratio of primary to secondary air during combustion—primary air passes through the MR-750 burner, secondary air passes around the burner.

Placing firebricks on the lid to dampen the flue exit controls the draft and reduction. This kiln can be fired in full oxidation with a flue exit dampened to an 8 to 10 inch opening, or an area of approximately 23–28 square inches. Due to the drag caused by the walls in the flue, decreasing the flue area may negatively impact the draft, especially if converting an electric kiln that has large element grooves.

A 0-30 PSI pressure regulator and 0-15 PSI gauge are used to control the gas flow to the kiln. The shut-off valves on the burner pipes are left fully open and the dampeners on the Venturi burners are left at ¾ inch open and not adjusted. A reduction atmosphere is achieved by dampening the flue exit with soft bricks. For the 28×32-inch kiln and beginning at 1652°F (900°C), the flue exit is dampened to 2¼ inches to stall the firing and create heavy reduction. Dampener adjustments as little as plus and minus 1/8 inch are made to keep the temperature between 1652°F (900°C) and 1742°F (950°C) for around 45 minutes. Since both the indicated pressure and the size of the dampened flue exit can be recorded in a log, there is direct feedback to help learn the best firing settings.

Make sure that the space between the bottom shelf and kiln wall leading to the flue contains at least 25 square inches for a 28-inch-diameter kiln and 18 square inches for a 27-inch-diameter kiln. This ensures that there won’t be a draft restriction leading down around the bottom shelf to the flue entrance. This also applies to any baffle or supporting blocks under the first shelf. And lastly, the flue exit in the lid should be on the side of the kiln, not at the back next to the hinge. I found that placing the flue exit near the hinge weakens the area and bricks will crack when the lid is lifted repeatedly. Also, the spy hole must be on the side and not in the area above the burners.   

I am really pleased with how easily the kiln fires, the amazing reduction, and the repeatability of results. Utilizing thin kiln shelves as a thermal draft inducer proved to be the key in creating this inexpensive and simple-construction design, and I hope that it will provide more potters with access to cone-10 reduction firing. For a more detailed description of the conversions, including additional photos, dimensions, and firing schedules, visit www.sebastianmarkblog.com/2018/07/gas-kiln-conversion-downdraft.html.

Safety

The converted kilns were all placed in a steel shed and a hood connected to 6-inch steel ducting vents the kilns to the outside. As I do not leave the kilns unattended during the 6 hours or so of firing, I did not install a Baso safety valve. Simple pilot lights are used to ignite the burners and are not needed once the kiln gets red hot. However, some locations may require a Baso system per code, especially if the kiln is attached to a household propane tank. I do not candle overnight, but instead have an initial 10–15-minute period where the burner is set very low and the lid is open about ½ inch, and then a slow warmup continues for around 15 more minutes. This pre-heating has been sufficient to warm things up and get enough heat flowing so the draft works once the lid is closed.

the author Boris Robinson’s education was in both the arts and the sciences. The two have intermixed throughout his life, from inventing and patenting a dry-gas mass flowmeter in the 1980s, to working with black-and-white photography and printing with carbon gelatin, and more recently to making Tokoname teapots.

1 One such online resource for calculating flue specifications is https://www.engineeringtoolbox.com/natural-draught-ventilation-d_122.html.