My Candle Develops a Complete Air Gap Around Sections of the Glass — Should I Change Pouring Temperature Before Changing Wax?

What a complete air gap is — and what it is not

Start here, because the name misleads everybody. The trade calls these marks "wet spots", and the word sends makers hunting for water, condensation, monsoon humidity or fragrance oil leaching out of the wax. What is actually there is air. Soy wax contracts as it goes from liquid to solid, and it keeps responding to temperature after it has set. Where it pulls off the glass it leaves a thin cavity, and light passing through the glass meets air instead of wax. Air and wax bend light differently, so the cavity reads as a darker, damp-looking shadow with a sharp edge. Run a finger down the outside of the jar and it is dry. Tip the jar and nothing moves. For the full version of this argument, see why does my candle have wet spots and the sibling post on whether the wax became wet or pulled away from the jar.

Your particular symptom — a complete gap around whole sections of the wall, not freckles and not a thin band — tells you one useful thing immediately. A continuous release over a large area is a bulk behaviour. It is the whole body of wax moving as one, which points at the pour and the cooling and at how the glass was prepared, rather than at something that happened in one small place. A pattern of scattered specks points somewhere else entirely, and a gap confined to the middle third of the jar points somewhere else again. Those are different posts for good reason: in this series the pattern is the diagnostic.

The second thing to be clear about before you spend anything: a gap, by itself, is cosmetic. It does not make the candle unsafe and it does not change the burn. The melt pool forms at the top and works down; a void against the glass three centimetres below the surface is not involved. What it does is cost you money at the point of sale, because a customer reading a clear jar sees damp, or oil, or carelessness. That is a real commercial problem and it deserves real effort — but it is an appearance standard you set, not a fault that fails a candle. The genuine safety exceptions stay exceptions: cracked glass, a wick that has drifted against the wall, a wick that clogs and smokes, or a jar too hot to hold. Those candles do not ship whatever the wax looks like.

The cost ladder: six levers, cheapest first

Here is the whole decision on one table. The column that matters is not "how likely is this" — nobody can tell you that from a description, and anyone who does is guessing. The column that matters is what it costs you to find out. Work down the list and stop when the gap closes. Most workshops never reach row six.

Two honest notes on that table. First, the ordering is economic, not physical — row six is not the least likely cause, it is the most expensive one to investigate. Second, rows one to four are not just cheap, they are reversible. If warming the glass does nothing you stop warming the glass and you have lost ₹531 of hardware that still earns its keep smoothing tops. If a new wax does nothing you have 10 kg of it, a recipe that no longer matches your documented temperatures, and a wick size you have to re-prove.

What pour temperature can and cannot do

Pour temperature is the lever everyone reaches for, and it is worth understanding why it is plausible without overclaiming it. When molten wax meets glass, the glass takes heat out of the wax at the contact face. The wax at the wall therefore sets before the wax in the middle, and it sets while the bulk behind it is still shrinking. That is the mechanism by which a wall-contact problem and a contraction problem are the same problem seen from two sides. Pour warmer and the wall-side wax stays mobile a little longer. Pour cooler and it sets sooner.

So far, so good. Here is the part the forums leave out: hotter is not simply better. A hotter pour is also the classic route to a rough, cratered or sunken top, to a bigger cavity around the wick, and to more visible crystal texture after cure. You can absolutely close a gap by pouring hotter and finish with a candle that fails inspection for a different reason. That is why the answer is never "pour at 85 instead of 75". It is a ladder of two or three degrees at a time, scored on both faces of the candle — the wall through the glass, and the top from above. Related reading: wet spots and sinkholes — the pouring temperature truth and why candle wax cracks on top.

Your ladder has to live inside what your wax page actually publishes. Three CSI waxes publish a pour temperature and the rest do not — and a silent page is information too, because it means the window is yours to establish and record rather than to look up.

Notice the jump between CSI 464 and CSI 468: a thirty-degree difference in published pour temperature between two container waxes in the same range. That is the single best argument for doing the free experiment first. A pour temperature that is near the bottom of one wax's window would be far below another's, so "pour hotter" is only meaningful relative to the wax in your pot. And a 150 g clear jar has its own ceiling: the Clear Glass Jar with Golden Lid page states it is safe for hot pour up to 85°C, which caps any ladder you run in that vessel regardless of what the wax permits.

The 12-jar paired pour-temperature test

This is the experiment that earns the right to spend money later. It uses one melt, one wax, one oil, one wick size and one jar, and it changes exactly one thing: the temperature at which the wax leaves the jug. Twelve jars is the smallest number that gives you six of each and still lets one or two be odd without wrecking the comparison. Every jar is sellable if it passes, which is why the marginal cost of the test is nothing.

The three things worth buying while you work down the ladder

None of these is a fix you can buy. Two of them are what the free experiments need in order to be experiments at all, and the third is the cheapest material change available — which is to say, the first row on the ladder that actually costs money. Each card shows every listed pack with the per-unit maths and how many 150 g candles at 8% a pack covers.

When the wax change is the right call

Changing wax is a legitimate answer. It is just an answer with a price tag and a date on it, and the point of the ladder is to make sure you buy it for a reason. You have earned row six when all of the following are true: the glass is cleaned and handled consistently, the pour temperature has been laddered across the full stated window with no improvement at the wall, warming the glass made no difference you could photograph, and the gap is still large enough that you are rejecting candles you would otherwise sell.

Then there is the honest question of what you are buying. Two CSI wax pages do use the phrase "excellent glass adhesion" — CSI 464 ("Smooth, opaque finish with excellent glass adhesion", "a creamy appearance, and reduced frosting compared to conventional soy waxes", "Low frosting, minimal blooming") and CSI 468 ("Formulated for excellent glass adhesion and a one-pour finish", "Rich, opaque appearance with minimal frosting", an "ultra-smooth finish"). Read those for what they are: qualitative product descriptions with no measurement, no scale, no test method and no comparison figure behind them. There is no published adhesion number anywhere in the range you could put in a spreadsheet, which is exactly why a wax change still has to be qualified in your jar, with your oil, at your load. The claim tells you where to look. Your twelve jars tell you what happened. Several pages — Natural Soy Pearl Wax, Premium Coconut Soy Wax, Coconut Wax — say nothing about appearance at all, and that silence is worth noting rather than filling in.

That last row is the one makers underestimate. CSI 468's stated pour temperature is 58–60°C against Luxury Soy Wax Chunks' stated 70–80°C, and its stated melting point is 50–53°C. Switching wax does not just change a line on your purchase order; it changes the temperature your team works at, the moment they pour, how long they wait, and every number on the batch card. If the free pour-temperature ladder had worked, you would have kept all of that. For the wider wax-selection argument, see best wax for jar candles in India and the sibling post on diagnosing wet spots before replacing raw materials.

The one-line record that stops you repeating this

Everything above collapses if you cannot read your own notes in three weeks. One line per jar, written at pour and completed at each photograph, is enough — and it is what turns "I think hotter helped" into a pour temperature you can put in an SOP and hand to someone else.

Record, in this order: date · wax and pack · wax lot or bag · oil and grams · load % · jug temperature at pour · glass state (room, warmed, cleaned how) · jar handle and pack · wick code · room temperature · cooling position · wall score at 2 h, 24 h, day 7 · top score at 24 h and day 7 · verdict. Fourteen fields, about twenty seconds a jar. For the comparison-recording method in more depth, the sibling post on what to record when comparing glass adhesion across waxes goes field by field.

Two habits make the record trustworthy. First, score before you look at the log — if you know jar H4 was the hot one, you will see what you expect to see. Have someone else read the scores off the photographs with the labels covered, or at least score from the photographs rather than from the jars on the bench. Second, keep one jar from every batch forever, labelled and lidded, on a shelf away from sunlight. A twelve-month shelf of reference jars is the cheapest quality system a candle workshop can own, and it is the only way to tell next March whether this March's wax lot behaves like last year's.

Frequently asked questions