How the kiln energy calculator works

A kiln at temperature leaks heat through its walls at a steady rate. That rate is what the elements or the burner have to replace just to hold, before you have heated a single pot. The rate sets the power you need. The energy calculator works it out the honest way, as a coupled steady-state balance rather than a single R-value division, then turns that number into recommended power and a fuel rate.

This post shows how to use the tool, then teaches the physics behind each number so you know what you are reading.

How to use it

You give the tool the kiln and the firing. Enter the interior dimensions of the chamber (width, depth, height) in inches or centimetres. Pick the brick grade, K-23 or K-26, and the wall thickness. Add your backup fibre: the number of blanket layers, the thickness per layer, and the density grade from 64 to 160 kg/m³. Set the target cone. Choose the firing method: electric, natural gas, propane, or wood. Then describe the firing: the hours to climb to temperature, the hours of soak at the top, the mass of ware and shelves in the load (or let the tool estimate it from the chamber volume), and what you pay per kilowatt-hour or per unit of fuel.

The tool reads back the hold and the firing. For the hold: the heat loss at temperature, the R-values of the brick and the fibre, the exterior shell temperature flagged as safe, hot, or a burn hazard, and the recommended power, shown as element kilowatts for electric or a burner rate for combustion. For the firing: the energy to heat the kiln’s own mass, the energy to heat the ware and shelves, the loss that leaks out across the whole schedule, their total, the fuel that total works out to after efficiency, and what it costs at your price. Change any input and every number updates, so you can watch the shell cool as you add a fibre layer, or watch the cost climb as you move up a cone.

The steady-state balance

Heat conducts from the hot face outward through the brick, then through the fibre backing, to the exterior shell. From the shell it leaves by two paths: natural convection into the surrounding air, and radiation to everything cooler than the shell. At steady state those flows are equal. Whatever crosses the wall must leave the surface. The shell settles at whatever temperature makes the two sides match.

Q = (T_int − T_s) / R_cond = h·A_ext·(T_s − T_a) + εσA_ext(T_s⁴ − T_a⁴)

The left side is conduction through the wall. The right side is convection plus radiation off the shell. Only one shell temperature satisfies both. The tool solves for it with a damped Newton-Raphson iteration. That coupling is the whole point: the conduction depends on the shell temperature, and so do the two surface losses, so you cannot solve one without the others. Ambient is fixed at 25 °C. Emissivity is 0.9, typical of painted or oxidised steel. The radiation term is fourth-power in temperature, which is why a bare hot shell sheds heat fast, and why the tool flags any shell over 70 °C as a burn hazard.

Conductivity is not a constant

Insulating firebrick and ceramic fibre both conduct more heat as they get hotter, often two to three times more at cone 6 than at room temperature. K-23 brick reads about 0.13 W/mK at 260 °C and about 0.27 W/mK at 1260 °C. Plug a single catalogue k into the R-value and you badly underestimate the loss.

So the solver does not use one number. It reads k(T) for each layer at that layer’s mean temperature, recomputes the R-values, re-solves the surface balance, and repeats until the temperatures and the conductivities both stop moving. Denser blanket conducts less at high temperature, which is why the density grade changes the answer. The brick curves are Morgan Thermal Ceramics’ K-IFB data (K-23 and K-26); the fibre curves are the Cerablanket data at 64 to 160 kg/m³.

Counting the wall: two methods

A thin flat panel loses heat through the same area on both faces. A kiln wall does not. The wall is thick relative to the chamber, so the heat-flow area grows from the small interior face to the larger exterior face. Conduct through the interior area alone and you count too little wall; conduct through the exterior area and you count too much. The tool offers two ways to settle this, and shows the figure from whichever one you did not pick as a cross-check.

Mean area conducts each layer through the average of its inner and outer face areas. The brick uses the mean of the interior area and the brick-to-fibre interface area; the fibre uses the mean of the interface area and the exterior area. It is simple, it is the conservative choice, and it credits the material at the twelve edges and eight corners of the box as if heat went straight through it.

A_brick = (A_int + A_iface) / 2

Shape factor is the furnace designer’s method, from Langmuir’s measurements on thick-walled furnaces and tabulated in Holman’s Heat Transfer. Heat does not go straight through an edge: the flux lines fan out over a longer path, so edge material conducts less than its face area suggests, and corner material less again. Each layer gets a conductance S built from its flat inner faces at full value, its twelve inner edges at 0.54 per unit length, and its eight corners at 0.15 times the layer thickness. The loss is then k · S · ΔT.

S = ΣA_inner / L + 0.54 · Σ edges + 8 · 0.15 · L

For a cube of inner side a and wall L, mean area gives 6a²/L + 12a + 12L and the shape factor gives 6a²/L + 6.48a + 1.2L. The flat-face term is the same. Mean area counts the edges about twice and the corners about ten times too heavily. On a big kiln with a thin wall the flat term dominates and the two agree within a few percent. On a small studio kiln, where the wall is a third of the chamber, mean area lands 15 to 25% high. The shape factor is the default; switch to mean area when you want the conservative figure for sizing elements rather than the closer estimate of running cost and shell temperature. Neither is exact. Langmuir’s factors are themselves good to a few percent, and a real kiln has a door and a floor that are not quite the same wall as the sides. Surface loss always uses the true exterior area, because that is the real surface the room sees.

You can see all of this in the tool. The wall cross-section draws the real temperature curve through the layers. The R-value stack-up shows how the total resistance splits between brick and fibre. Drop the fibre to zero layers and both the shell temperature and the loss jump.

From loss to power

Steady-state loss is not the size of kiln you buy. It is the minimum to hold temperature. Multiply the loss by a 1.25 hold margin and you have the smallest element package that could, in principle, sit at cone 6 forever. To actually climb a firing schedule you also have to heat the brick mass and the ware from cold, so a full firing needs several times the hold loss. The tool assumes roughly three to five times.

To keep the sizing honest, the tool shows the physics figure next to four rule-of-thumb methods drawn from practice:

• the survey medians from our 60-kiln survey, 67.6 W/L and 0.635 W/cm²;
• Olsen’s volume rule, 61 to 73 W/L (1.0 to 1.2 W/in³);
• Olsen’s surface rule, 0.78 to 1.09 W/cm² (5 to 7 W/in²).

When the physics minimum and the shop-floor rules bracket each other, you can size with confidence. When they diverge, the design is usually unusually well or poorly insulated, which is worth knowing before you order elements.

Fuel rate

For a combustion kiln the input rate is the heat loss divided by a system efficiency, because much of the flame’s energy goes straight up the flue. The tool uses about 100% for electric resistance, 50% for natural gas and propane, and 20% for wood. It then converts the gross input to a fuel rate using each fuel’s higher heating value: 37.3 MJ/m³ for natural gas, 50.4 MJ/kg for propane, and 20 MJ/kg for dry wood. With inches, gas reads in therms and cubic feet, propane in gallons, and wood in pounds priced by the 2500 lb cord; with centimetres the same figures come out in cubic metres and kilograms. With gas or wood the heat rates read in BTU/hr rather than watts. Treat those as ballpark hold rates, not billing figures. Real efficiency depends on your burner, your damper, and how you fire.

What it does not do

The tool models a clean box with uniform walls, heated on a straight ramp to a fixed temperature and held there. The ramp is taken as linear in time, and the loss during the climb is the steady loss at each temperature along the way. A real wall lags behind the chamber on the way up, so that ramp loss runs a little high, and the stored heat is charged at the wall’s final gradient. Cooling is not counted; once the power is off nothing more is spent. Ambient is fixed at 25 °C, so a cold winter studio or a hot kiln room shifts the real numbers. The insulation is taken as fresh; an old K-23 wall has gained density from cycling and loses more. Nothing prices out door gaps, peep holes, flue losses, or a fibre lid over a brick body, and the heat absorbed by the reactions in the clay itself is left out. Read the figures as a floor, not a bill.

Sources

• Morgan Thermal Ceramics. K-IFB insulating firebrick data (ASTM C-182 hot-wire k(T)) and Cerablanket blanket data (ASTM C-201), 64 to 160 kg/m³. Source for the temperature-dependent conductivity tables in the solver’s outer loop.
• Holman, J. P. Heat Transfer, Table 3-1 (conduction shape factors), after Langmuir’s furnace-wall measurements. Source for the edge (0.54) and corner (0.15·L) factors in the shape-factor method.
• Olsen, Frederick L. The Kiln Book, Chapter 9. Source for the power rules of thumb (volume 1.0 to 1.2 W/in³, surface 5 to 7 W/in²) and the gas burner sizing rule, 10,000 BTU/ft³·hr for insulating firebrick.
• What 60 electric kilns tell you about power. The survey behind the 67.6 W/L and 0.635 W/cm² medians the tool sizes against.

Next step

Loss and power tell you how big the element package is. They do not tell you how to build it. Once you have a target wattage, the element calculator turns it into a real coil: wire gauge, alloy, resistance, coil length, and the groove it has to sit in. Read how the element calculator works for the wiring physics, then carry your interior dimensions straight across with the tool’s Size the elements button.