Sinkholes vs Cracks vs Air Pockets — How to Diagnose Candle Wax Problems

Tell the three apart, precisely

Before you change a single thing about your process, name the defect correctly. This matters more than it sounds, because the three respond to opposite adjustments. Slowing the cool-down fixes cracks and shrinks sinkholes; it does nothing whatsoever for air whipped in during stirring. Pouring cooler reduces contraction voids; poured too cool, it puts more bubbles in, not fewer. If you are treating the wrong defect you are not standing still, you are going backwards.

Where it appears: almost always around the wick, in the upper third of the candle, because that is the last region to freeze and the region with the most surrounding wax pulling on it. What it looks like: either an open crater in the surface, or — far more often, and far more annoying — nothing at all. Wax skins over the top while the interior is still liquid, and the void forms underneath a lid of perfectly smooth wax. The tactile tell: press the surface firmly with your thumb about 5 mm from the wick. Sound wax is unyielding. Wax over a void gives, flexes, or breaks through with a small dry crack. A second tell is weight — a jar with a large hidden void weighs noticeably less than its neighbours from the same batch, which is why a weighing scale at ₹354 doubles as a quality-control tool: weigh five jars from every batch, and anything more than a few grams light gets pressed and checked.

The physics of it — why liquid wax occupies more volume than the same wax solid, and why the shortfall collects where it does — is covered properly in why does my candle have a sinkhole. That is the page to read for the mechanism. This one is about telling it apart from its two neighbours and dealing with all three at once.

Where it appears: anywhere, but most commonly running down from the surface, or radiating out from the wick towards the glass. In pillars and moulded work it tends to run along the length. What it looks like: a clean line, often white or lighter than the surrounding wax because it is a fresh fracture face scattering light. It can be a hairline you only see at an angle, or an open split with visible width. The tactile tell: run a fingernail across it. A crack catches. A pour line or a stack mark does not. A crack in a container candle also tends to be audible when it forms — makers who pour in winter and hear a faint tick from the cooling rack are hearing wax fracture.

Cracks are a stress fault, not a volume fault. The wax has already set; then one part of it contracts faster than the part next to it, the stress exceeds what set soy will take, and it splits. That is why cracks cluster in cold-weather pours, in jars set on a cold stone or metal surface, in candles put into a fridge or in front of a fan, and in the batch that someone moved from a warm room to a cold one an hour after pouring. Soy Pillar Wax and moulded work are more prone than container candles simply because there is no glass wall holding the shape together.

Where it appears: against the glass, where you can see it, and scattered through the body where you cannot until the candle burns down to it. What it looks like: round or lens-shaped, with a smooth curved boundary. Against the wall it usually flattens into a disc — a bubble that rose, met the glass and stalled there as the wax set around it. In the body it is a small sphere, often in clusters. The tactile tell: there generally isn't one, because most of them are sealed inside. A surface pocket that has broken open leaves a small round pit with a smooth rim, which is the fastest way to distinguish it from a sinkhole — a sinkhole crater has a rough, irregular, drawn-down edge; an air-pocket pit is round and smooth.

The diagnostic table

One table, six rows, all three defects. Find your row, read across. The last column is the one most guides leave out and the one that should decide whether you re-pour a batch or ship it.

Why all three are the same problem

Here is the argument this page exists to make. Sinkholes, cracks and air pockets are three symptoms of one underlying discipline problem, which is thermal control. They present differently because they arise at different moments — the air pocket at the instant of pouring, the sinkhole during the phase change from liquid to solid, the crack afterwards in wax that has already set — but every one of them is governed by the temperature the wax was at, how fast it lost that heat, and how evenly.

This is genuinely good news for a business, because it means you are not chasing three fixes. You are tightening one process. A maker who buys a thermometer, writes down four numbers per batch and stops moving jars around while they set will usually watch all three defect types fall together over a fortnight. A maker who buys a different wax for the sinkholes and an additive for the cracks has spent money and changed two variables at once, and now cannot tell which — if either — did anything.

What hurts the burn, and what is only cosmetic

This is where a lot of troubleshooting advice quietly wastes your money, and it is worth being blunt about it. Most of what you are looking at is cosmetic. It affects whether the candle photographs well and whether a customer thinks it looks handmade or looks careless. It does not affect whether it burns properly. Knowing which is which decides whether a jar is a reject, a second, or a perfectly saleable candle you happen to be annoyed with.

A hidden void round the wick genuinely matters. It is the one on this page that changes how the candle behaves. When the flame burns down to a large cavity, the melt pool drops into it — you get a sudden deep pool where there should be a shallow one, the wick can drown in the extra liquid wax, or, if the void separates the wick from the surrounding wax, the flame runs out of fuel and starves. Either way the burn stops behaving predictably, and unpredictable is the one thing a candle you sold cannot be. Press-test the surface, and top up anything that gives.

A hairline crack almost never matters. Set wax with a fine fracture in it burns exactly like set wax without one; the melt pool closes over it within the first burn and it is gone. It is a presentation issue on the shelf and nothing more. Surface air pockets also almost never matter — a few bubbles against the glass are the most-photographed and least-important defect in candle making, and the melt pool erases them as it descends. There are two exceptions worth naming. A deep crack running the full depth can give molten wax a route to travel down, which in a badly cracked pillar means it runs out of the side; and an air pocket sitting directly against the wick behaves like a small sinkhole when the flame reaches it. Both are uncommon. Neither is a reason to treat every bubble as a failure.

The two-week record sheet

Everything above is a hypothesis until you have data. The reason most makers never solve cooling defects is not that the fix is difficult — it is that the fault appears hours after the decision that caused it, and nobody remembers by then what the wax was at when it went into the jar. Four fields per batch, written at the bench, will find the pattern inside a fortnight.

Two weeks of that — say eight to ten batches for a working maker — is enough. You are not looking for perfection, you are looking for correlation: which column moves when the defect column moves. Almost always it is one of the four controls, and almost always it is obvious once it is written down in the same place. Then change one thing, run another five batches, and compare. If you change the pour temperature and the cooling surface in the same week and the defects go, you have learned nothing you can repeat. The same discipline applies to every other part of costing and quality in this business, which is why it shows up again in scaling to 1,000 orders a month.

The tools that actually fix it

Two purchases cover all four controls, and together they cost less than a kilogram of most fragrance oils. Prices are CSI live pricing, September 2026, and the linked product page is always authoritative.

Beyond those two, the only other spend worth considering is a proper pouring vessel, because pouring down the side of a jar with a good spout is a different job from pouring with a saucepan. The Candle Wax Pouring Jug is ₹531 and the 600 ml Steel Pouring Pitcher is ₹708. Neither is urgent. A thermometer, a notebook and a cooling spot that does not change will take you further than either, and none of it requires you to change the CSI 464 at ₹260 per 500 g or the Luxury Soy Wax Chunks at ₹299 per 500 g you are already using. If you are still building your bench, the full list is in the complete candle making raw material list, and wick sizing — which is a separate variable from everything on this page, and worth controlling before you start testing temperature — is covered in the wick guide by diameter. Eco Candle Wicks start at ₹59 for 10 if you need to hold the wick constant while you test.

Frequently asked questions