You’ve dialed the wax, sized the wick, cured for three weeks, and the burn test on your sample jar looks clean. Then the same recipe lands in a cheaper container and the first burn goes to soot, the melt pool stalls half an inch from the wall, and a customer photos a hairline crack near the base. The wax is identical. The wick is identical. The variable you never put on the spec sheet is the glass.
The vessel is part of the burn system, not packaging
Every candle flame runs at 1,300–1,400°F at the wick tip, while the liquid wax pool sits around 160–180°F. Heat has to get somewhere, and the glass wall is the biggest heat sink on the table. A jar that is 3/16 inch thick soaks up a very different amount of heat than one that is 1/8 inch, and that difference reaches back into the flame. The wick must supply enough heat to (a) vaporize fragrance, (b) sustain the melt pool, and (c) lose some energy to the container wall. When the wall gets thinner or the glass gets cheaper, the third term grows and the flame has to work harder to keep the first two stable. That is why a wick that runs clean in a heavy jar can over-run and black out in a light one.
What most burn charts miss is that the container is a load on the wick, not a passive sleeve.
Two glass families, two burn behaviors
Soda-lime glass, the standard for most candle jars, has a coefficient of thermal expansion around 9.2 ppm/°C and begins to soften near 1,000°F. Borosilicate, the same family as lab glass, sits at roughly 3.3 ppm/°C and softens closer to 1,400°F. In practice, that expansion difference is what separates a jar that shrugs off a 190°F pour into a cold vessel from one that develops a ring crack at the base within a few cycles. If you are sourcing at scale, ask your container supplier for the material type and the annealing temperature. If they will not answer, assume soda-lime and plan your pour protocol around it.
A second number matters: the wall’s own surface temperature during burn. In an 8 oz vessel, a full melt pool with a healthy wick will push the glass surface in the mid-200s °F within an hour. That number decides whether a thin jar warps a label, sweats condensation on the outside wall, or — in a sealed jar with a tight lid — builds enough internal pressure to push the lid off or dome the wax. The lid is not decoration. A tight-fitting lid in a thin-walled vessel is a slow thermal trap: it holds heat, slows the wax’s ability to off-gas fragrance during a burn, and accelerates frosting on the next cool-down.
The three failure modes that trace back to the vessel
1. Early soot and a short, hot flame
Thin glass steals heat faster. The wick compensates by burning taller, the flame zone gets narrower, and incomplete combustion produces visible soot on the jar wall. If a wick that runs clean in a 3/16 inch jar soots in a 1/8 inch jar, the fix is not a smaller wick — that starves the melt pool. The fix is a thicker wall, or a slower pour that lets wax and glass reach thermal equilibrium before the first burn.
2. The melt pool that will not reach the wall
A heavy, thick-walled jar soaks more heat per pass than a thin one. In an 8 oz vessel the difference between a 3/16 inch and a 1/8 inch wall is a meaningful fraction of the heat a 10-12 wick can deliver. The melt pool stalls, the wax behind it stays solid, and you get the classic tunnel even though the wick is correctly sized for the wax. This is the failure most often misdiagnosed as a wick problem.
3. The lid that becomes a pressure problem
A tight lid on a thin jar during a full burn traps heat and fragrance vapor. The wax off-gasses, the headspace warms, and the internal pressure rises. On cool-down the same headspace contracts and pulls moisture in. That cycle is the engine behind frosting and ghost scent in sealed vessels. Loosen the lid, or switch to a vented design, and both symptoms drop. If you are selling with a lid, the spec is not “snug” — it is “stays on the shelf, comes off with a light lift.”
A working spec sheet for the vessel
Put these on your purchase order and your burn log:
- Wall thickness at the equator — measure with a caliper, not from a data sheet. 1/8 inch is the floor; 3/16 inch is the working target for anything 8 oz and up.
- Material — soda-lime or borosilicate. If the supplier will not say, treat it as soda-lime.
- Base thickness — the base is the first place a thermal-shock crack starts. Under 3/16 inch at the base is a red flag for any vessel that will see repeated burns.
- Surface temperature at the first hour of burn — log it with an IR gun at the equator. Over 250°F in a thin-walled jar means wax and wick are carrying more heat load than the glass absorbs cleanly.
- Lid seal — should lift with a light press, not require a wedge. A lid that stays put under its own weight on a hot vessel is a pressure risk.
What to do before you commit to a vessel at scale
Run one vessel comparison before you order a thousand. Take the same wax, the same wick, the same fragrance load. Pour two jars — one from your current supplier, one from the new one. Burn both for three sessions, three hours each. Log the melt-pool diameter after hour 1 and hour 3, surface temperature at the equator after hour 3, and any soot on the inner wall. If the new jar’s melt pool is more than half an inch smaller at hour 3, the glass is stealing heat you cannot recover with the wick you are using. If the surface temperature is more than 20°F higher, the lid and the label are at risk. If the base shows any clouding after three cycles, it is a structural problem, not an optical one.
The vessel is not the container. It is the fourth variable in your burn equation, and the one most often left off the sheet. Put it on. The soot, the tunnel, and the cracked base all trace back to it.
At YS Candle, we spec wall thickness and lid clearance the same way we spec wick and wax load — because a burn test that does not name the vessel is a burn test that will fail in the field.