Kiln cooling schedules explained

A kiln cooling schedule is a set of programmed downward segments, the same idea as a ramp but in reverse. Controlled cooling does three jobs: it grows the crystals that make matte and satin surfaces, it heals late glaze defects, and it carries ware through the silica inversions slowly enough to avoid dunting.

Publie le 22 juillet 2026 · Source : Ceramic Arts Network, Super Cool: Slow Cooling in an Electric Kiln

Most firings end at the top of the climb. The controller shuts off, the kiln free-falls, and whatever happens on the way down happens by accident. That is fine for plenty of work. It is also the most overlooked half of a firing schedule, and the half that decides the surface of a mid-fire glaze.

Cooling is really three problems in three separate temperature bands, and each one wants a different treatment. Knowing which band you care about tells you which segment to write.

What is a controlled cooling schedule?

It is one or more programmed segments that hold the kiln back on the way down instead of letting it fall. A cooling segment has a rate in degrees per hour and a target, the same as a heating ramp. The controller holds that rate by pulsing the elements, which is why a kiln can cool slower than natural but never faster.

That last point matters more than it sounds. The fastest cool available is whatever the kiln loses on its own with the power off. A heavy, well insulated kiln packed with shelves cools slowly with no program at all, while a small thin-walled kiln drops fast. The same written program therefore does different things in different kilns, and your own firing log is the only reliable guide.

Why is free-fall cooling not always right?

Because a fast cool freezes the glaze in whatever state it was in at shutoff. Crystals never get time to form, late bubbles never heal, and the ware crosses the silica inversions at whatever rate the insulation allows. For glossy glazes on thin work that is often fine. For matte surfaces or thick sculpture it is not.

Ceramic Arts Network puts the relationship simply: a fast cool results in a glossy surface, while a slow cool goes matte. That one sentence explains a lot of frustrated testing. If a glaze that looked satin in the supplier photo comes out glassy in your kiln, the recipe may be perfectly fine and the cooling may be the whole difference.

The other cost of a free fall is stress. Ware is still shrinking as it cools, and it shrinks unevenly if the outside cools much faster than the inside. Thick walls, wide flat forms, and closed shapes pay for that first, usually as a crack discovered on unloading.

Where do crystals grow during cooling?

In a window below the peak, roughly 1900 down to 1450 F for the micro-crystals that make satin and matte surfaces at mid fire. Above that band the melt is too fluid for crystals to hold together. Below it the glaze is too stiff for anything to move. That band is where the surface gets decided.

Practical cooling rates through the window run from about 125 to 175 F per hour depending on how matte you want the result, with slower giving more crystal growth and a softer surface. Some potters let the kiln free-fall the first 300 F below peak, which drops it out of the very fluid zone quickly, then take control for the crystal band.

Large, showy crystals are a different project. Zinc silicate crystalline glazes want a hold of several hours somewhere around 1850 to 2050 F, which is a specialist firing with its own catch dishes and its own risk to shelves. The slow cooling for glaze effects guide covers the mid-fire version, which is the one most studios will use.

One caution worth repeating from the source: changing the surface changes the glaze. Test any new cooling program on tiles before committing functional ware, and leach test any result you plan to eat off.

How do I cross quartz inversion safely on the way down?

Slowly, and twice. Quartz inverts near 1063 F (573 C) with a sudden volume change, and ware that has formed cristobalite inverts again in the 410 to 536 F (210 to 280 C) band, most often cited near 473 F (245 C), with about double the volume change. A cooling rate of roughly 100 to 200 F per hour across both bands prevents most dunting.

Cristobalite forms mainly above 1922 F (1050 C), so it is a mid-fire and high-fire concern rather than a low-fire one. It is also why the tail of a cool down matters at temperatures that feel harmless. The kiln reads 500 F, the shelves look cold, and the ware is still crossing a volume change of around 7 percent.

This is where the temptation to crack the lid does real damage. Cool room air hitting one side of a pot at 400 F is exactly the shock the inversion cannot absorb. Keep the lid shut until the controller reads below about 250 F, and even then open it only to peek. Unload at room temperature.

What does a worked cooling schedule look like?

Three or four segments cover most needs: a fast drop out of the fluid zone, a controlled crawl through the crystal band, and a gentle pass across the inversions. The cone 6 example below is written as rates and targets. Treat it as a starting point to test in your own kiln, not a program to trust on a full load of finished work.

1. Segment 1: free fall, or about 500 F per hour, from peak down to 1900 F. This drops the glaze out of its most fluid state so it stops running.
2. Segment 2: 125 to 175 F per hour from 1900 F down to 1450 F. This is the crystal band. Slower gives a more matte surface.
3. Segment 3: about 150 F per hour from 1450 F to 950 F. This carries the ware through quartz inversion at 1063 F without a jolt.
4. Segment 4: off, or about 200 F per hour if your kiln is thin walled, from 950 F down. Keep the lid closed through the cristobalite band, roughly 410 to 536 F.
5. Unload at room temperature. Nothing good comes of opening early.

A drop-and-hold for pinholes sits above that sequence rather than inside it. After the cone bends, drop about 100 F below peak and hold 20 to 60 minutes, then begin segment 1 from there. Adding hours at high temperature does cost element life, so use the hold you need and no more.

What is a cooling hold, and when do you use one?

A cooling hold parks the kiln at a fixed temperature on the way down instead of continuing to fall. The two common uses are healing glaze defects just below peak, and growing crystals inside the matte window. A hold does more work than a slow ramp at the same temperature, because the kiln stays put rather than passing through.

The drop-and-hold is the best known version. After the cone bends, the kiln drops roughly 100 F below peak and holds there for 20 to 60 minutes. Fewer new gas bubbles are being generated at that temperature, while the glaze is still soft enough to close over the craters left by the ones that already burst, so pinholes and blisters often disappear where a top hold alone did nothing.

Crystalline holds sit lower and last longer. A hold of fifteen minutes about 60 F below the shutoff temperature is a common way to settle the melt before a controlled cool begins, and true crystalline glazes hold for hours around 1850 to 2050 F. Every hold adds heat-work, so a cone that used to land cleanly may over-fire once you introduce one. Read the witness cone after the first firing with a new hold in it.

Loading changes what a cooling program actually does. A kiln packed with heavy shelves and thick pots holds heat and cools slowly by itself, which means a programmed cool of 150 F per hour may be barely slower than the natural rate. An almost empty kiln, or one with thin walls, drops fast and needs the program to do real work. Log the actual times against the program, and you will learn quickly which of your firings need the segments at all.

Does slow cooling cost anything?

Time and elements. A controlled cool can add two to five hours to a firing, and it keeps the elements working at high temperature for longer, which is exactly what shortens their life. That is a real trade, worth making deliberately rather than leaving a slow cool switched on for every program out of habit.

Set the peak and the speed you want in the firing schedule builder, add a cooling ramp, and print the sheet. Then read a witness cone at three heights and see whether the kiln did what the program said. That comparison, repeated over a few firings, is what turns a published cooling schedule into your cooling schedule.