Annealing Schedules for Glass: Anneal Points, Soak Times, and Kiln Cycles by Glass Family

Short answer: Annealing means soaking finished glass at its anneal point — roughly 940–970 °F (505–520 °C) for COE 104 soft glass, ~950 °F (510 °C) for COE 96, 900 °F (482 °C) for Bullseye COE 90, ~1050 °F (~565 °C) for boro 33, and around 1140 °C for fused quartz — long enough for the whole piece to equalize, then cooling slowly through the strain point so no new stress is locked in. Soak time scales with thickness (from half an hour per centimeter for soft glass to an hour per quarter inch for boro), and thicker work needs slower cool-downs. Skip it and the stress stays in the glass, waiting — pieces can crack days, weeks, or months later. Figures vary by manufacturer and source, so treat everything below as a starting point and confirm the schedule for your specific glass.

Annealing is a kiln topic, but it’s inseparable from your torch decisions: the glass family you work — covered in soft glass vs boro vs quartz — sets both the flame you need and the kiln cycle your finished pieces demand. If you’re still picking that family, start with the complete glass torch buyer’s guide.

What annealing is, and why stress forms in the first place

When you pull a piece out of the flame, the outside cools and stiffens before the inside does. The surface locks into its final dimensions while the interior is still contracting; as the core catches up, it pulls against a shell that can no longer move. That mismatch is frozen into the piece as internal stress. Source: Hot Shot Ovens.

Annealing erases that stress. The piece is held at a temperature where the glass is still rigid but its structure can slowly relax — the anneal point — until the whole piece is at one uniform temperature and the stress has crept out of it. Then it’s cooled slowly through the temperature band where new stress can form, and only below that band can the cooling speed up.

What happens if you skip it? Sometimes nothing — at first. Unannealed work that survives the initial cool-down is still carrying its stress, and it can let go days, weeks, months, or even years later: beads split in half, sculptural pieces drop appendages, and any extra insult (a knock, a temperature swing, later coldworking) becomes the trigger. Unannealed glass is simply weaker and less thermally stable than the same piece properly annealed. Source: Hot Shot Ovens.

Anneal point vs strain point (the two temperatures that define the cycle)

Every glass has a characteristic pair of temperatures that bracket the annealing process:

• Anneal point — the temperature at which stress relaxes in minutes. This is where you soak: the glass is hard and holds its shape, but internally it can still rearrange enough to release tension.
• Strain point — the lower temperature below which the glass is fully rigid and no new stress can be introduced (or relieved). The zone between anneal and strain point is the danger band: cool through it too fast and you re-create the stress you just soaked out.

So every schedule has the same skeleton: soak at the anneal point → controlled slow ramp down to below the strain point → faster cool to room temperature. The soak length and the ramp speed are what change with glass type and thickness. Sources: Kiln Frog, Northstar Glassworks.

Anneal and strain points by glass family

Figures below are typical published values, rounded; sources genuinely disagree by 10–30 degrees on some of these, so always prefer the number from your glass’s manufacturer.

Sources: Lampwork Etc. COE 104 poll, Kiln Frog, Northstar Glassworks, Kavalierglass Simax TDS, Wikipedia — Fused quartz.

Quartz deserves its own caveat: with an anneal point around 1140 °C, it’s outside the range of ordinary studio annealers entirely — one of several reasons it stays a specialist material, as covered in soft glass vs boro vs quartz.

Typical schedules: soak time scales with thickness

The soak exists so the entire cross-section reaches anneal temperature and equalizes — so thicker work needs longer soaks and, just as important, slower ramps down. Common published rules of thumb:

Sources: Lampwork Etc., Kiln Frog, Northstar Glassworks.

Two practical notes. First, thickness means the thickest part, not the average. Second, the community consensus is that over-annealing is not a real risk at these temperatures — when in doubt, soak longer and cool slower. The failure mode is always too little, not too much. Source: Lampwork Etc..

Garaging: keeping work hot during a session

Garaging is not annealing — it’s the holding pattern before annealing. During a session, each finished piece goes into a hot kiln held above the strain point, at or slightly below the anneal temperature (bead makers with kiln doors often garage a few tens of degrees under the anneal setpoint), so it never cools into the stress zone while you keep working. When the session ends, you run the actual annealing cycle on everything at once. Source: Chaotic Glass — Mike Aurelius.

Garaging is what makes larger work possible at all: a big bead or a thick boro piece set on the bench to cool will often crack before you ever get to anneal it. A kiln radiating at 950 °F+ within arm’s reach of the torch also belongs in the same planning exercise as ventilation and gas routing — see the glass torch safety and setup guide.

Batch annealing vs annealing as you go

There are two honest workflows:

• Annealing as you go (garage-then-anneal): every piece goes straight from the flame into a hot kiln, and the full cycle runs at the end of the session. This is the gold standard — the glass never passes through the stress zone uncontrolled.
• Batch annealing: pieces are flame-cooled, slowed in vermiculite or a fiber blanket, allowed to reach room temperature, and later loaded into a kiln, brought back up to the anneal point, soaked, and cooled properly. This works — small soft-glass beads usually survive the interim — but the unannealed pieces can crack before or during the reheat, and larger or more complex work often doesn’t survive the wait. It’s a budget-kiln strategy, not a superior one. Source: The Alchemistress.

Leaving pieces in the blanket and calling them done is not a third workflow — see the myths below. If you’re budgeting a first setup, this is why the kiln is part of the real cost of lampworking, right alongside the torch — see the best beginner glass torch guide for how the whole starter package fits together.

Checking your work: the polariscope

You can’t see stress in ordinary light, but you can see it in polarized light. A polariscope — two polarizing filters with the glass between them, plus a light source — exploits the fact that stressed glass becomes birefringent: stress splits light into two rays that travel at different speeds, and the recombined light shows up as bright fringes against the dark field. Well-annealed glass reads dark and quiet; stressed glass glows around the stress concentrations, which is exactly how a polariscope validates an annealing process. Source: Barnett Technical Services.

Clear boro is where this matters most for lampworkers: clear glass shows fringes plainly, and boro workers routinely polariscope test-pieces to validate a kiln schedule. Opaque and dense colors can’t be read this way — one more reason to trust the schedule rather than the absence of visible cracks.

Common myths

• “Vermiculite / a fiber blanket anneals my beads.” No. Vermiculite, fiber blankets, ash, and annealing bubbles only slow the cooling — they never hold the glass at its anneal point, so the stress stays. They’re a legitimate bridge to a later batch anneal, not a substitute for one, and for larger pieces they don’t slow cooling enough even for that. Source: The Alchemistress.
• “It didn’t crack, so it’s fine.” Surviving the first night proves nothing — stress can sit quietly and release much later. Uncracked and unannealed just means not cracked yet. Source: Hot Shot Ovens.
• “Small beads don’t need annealing.” They’re the most forgiving case, but they still carry stress, and anything sold, worn, or given away should be annealed.
• “More annealing can hurt the piece.” At proper anneal temperatures, over-soaking is not a practical concern — err long and slow. (Soak far above the anneal point and you have a different problem: slumping, not annealing.)
• “One schedule fits all glass.” Anneal points differ by hundreds of degrees across families — a boro schedule would slump soft glass, and a soft-glass schedule does nothing for boro. The schedule follows the glass, just as the torch does; see soft glass vs boro vs quartz.

Key takeaways

• Annealing = soak at the anneal point, then cool slowly through the strain point. The soak equalizes the piece; the slow ramp prevents new stress.
• Typical anneal soaks: ~940–968 °F for COE 104 soft glass, ~950–960 °F for COE 96, 900 °F for Bullseye COE 90, ~1050 °F (~560–565 °C) for boro 33, ~1140 °C for fused quartz — and sources vary, so confirm with your glass’s manufacturer.
• Soak time and ramp speed scale with the thickest part — rules of thumb run from ~30 min/cm (soft glass) to ~1 hr per 1/4” (boro), and thick work needs hours, not minutes.
• Garage during the session, then run one full cycle at the end.
• Batch annealing from cold works for small beads but adds risk; fiber blanket and vermiculite only slow cooling and are never annealing by themselves.
• A polariscope shows the stress you can’t see — test your schedule instead of trusting survival.

Sources

• Northstar Glassworks, “Borosilicate Glass Annealing Chart” — https://northstarglass.com/annealing/
• Kavalierglass, “Technical Data Sheet — SIMAX glass” — https://www.kavalier.cz/wp-content/uploads/2025/07/TDS-SIMAX-glass-2025.pdf
• Kiln Frog, “Recommended Annealing Cycle for Bullseye Glass” — https://kilnfrog.com/pages/recommended-annealing-cycle-for-bullseye-glass
• Wikipedia, “Fused quartz” — https://en.wikipedia.org/wiki/Fused_quartz
• Chaotic Glass (Mike Aurelius), “Garaging and annealing, two different aspects of the same beast” — https://mikeaurelius.wordpress.com/2008/01/30/garaging-and-annealing-two-different-aspects-of-the-same-beast/
• Hot Shot Ovens, “Inside Glass Annealing: What It Is, Why It Matters” — https://www.hotshotovens.com/blogs/glass-art-blog/annealing-glass
• Lampwork Etc., “POLL — annealing schedule for 104 COE” — https://www.lampworketc.com/forums/archive/index.php/t-83377.html
• Barnett Technical Services, “Strain Inspection in Glass with Polariscope Strain Viewers” — https://barnett-technical.com/strain-inspection-in-glass-with-polariscope-strain-viewers/

Editor’s note: anneal-point, strain-point, and schedule figures are typical published values as of 2026, rounded for comparison; they vary by manufacturer, specific glass formulation, and source. Follow the annealing chart from your glass’s manufacturer and your kiln maker’s instructions for the piece sizes you actually make.