Can a Candle Jar Survive Early Burn Tests but Still Crack When the Candle Burns Closer to the Bottom?

What an early burn test actually proves

Most makers' first burn tests are sensible: light the candle, time how long the pool takes to reach the edge, check the flame and the soot, repeat for a few sessions. If all is well, the jar is approved. The trouble is what those sessions cover. In our worked example jar, poured 6 cm deep, the first three or four sessions typically take the top third of the wax. During that time the flame sits at the top of a column of solid wax, and the only glass it can heat directly is the thin band of wall above the pool.

Here is what each stage of a candle's life can and cannot tell you, before any customer finds out for you.

Notice that the final third is also where the customer's experience ends. They light it on a weekday evening, burn it through a long Sunday, and finish it months after you poured it. If the jar is going to disappoint anyone, the final third is where conditions are hardest. Why that last stage matters commercially, for complaints and returns, is the final-third approval post; this one is about why the physics changes.

Five things that change as the candle burns lower

None of what follows is exotic. Each is a common cause of late-burn heat or late-burn cracking in candle glass, stated as a tendency, not a certainty. The diagram above shows the first three; the last two you only see by watching.

1. The wax buffer shrinks

Solid wax under and around the pool soaks up heat and spreads it. At the first burn, our worked example has 6 cm of wax below the flame. After the top third it has 4 cm; after the middle third, 2 cm; at a 1 cm stop line, 1 cm, which is 17% of what you poured. The same flame now has far less wax to warm before its heat reaches the glass base. That is the main reason many makers stop burns with about 1 cm left, and why many experienced makers approve a jar only after full burn-down tests.

2. The flame moves toward the base

At the stop line, the flame sits about 1 cm above the inside floor, directly over the thickest, slowest-to-heat part of many jars. Glass that heats unevenly (hot in one area, cooler around it) is under stress, and a sudden temperature difference is the textbook route to a crack, called thermal shock. If the wick tab is exposed or the pool is shallow, the base centre gets heat that the top of the candle never delivered. Jars that narrow toward the base or have a thick glass base make this stage matter more; see how a thick glass base changes the late burn.

3. The bare wall grows

Above the pool, the inside of the jar is bare glass, heated by the flame's hot air as it rises. At the first burn in our example, that band is 1 cm, or 14% of the inside depth. At the stop line it is 6 cm, 86% of the inside depth, six times taller. The jar is now a shelter around the flame, and a sheltered flame can burn taller and hotter than the same wick in the open. This is why a jar that only felt warm on the first burn can run noticeably hotter at the end, and why touching it is not a test (use your eyes and your log instead).

4. Wicks lean, carbon up and flare

After many hours, cotton wicks commonly grow carbon at the tip ("mushrooming"), lean as the wax softens around them, and burn with a taller, less steady flame. A wick leaning a few millimetres toward the wall matters little at 6 cm of wax; at 1 cm, in a jar only 6.3 cm across inside (our assumption), it can bring the flame close to the glass. A flame touching or nearly touching the glass is a common cause of cracks. Trimming before every burn and centring at the pour reduce this; they don't remove it.

5. Heat stacks up over a long session

Customers don't burn in tidy three-hour sessions. Late in a candle's life, the jar starts each session already carrying heat from the glass, and a long burn keeps adding to it. Debris makes it worse: wick trimmings, match ends or dust in the pool can act as extra fuel and flare. Draughts from a fan, an air-conditioner vent or an open window add uneven cooling on one side, which is another route to thermal shock. None of these show up in a short, early, supervised test.

Why a listing's heat wording cannot answer this

Of CSI's 72 container listings, only seven state a numeric heat figure, and all seven use the same words: "Safe for hot-pour up to 85°C" (the Clear 150 g jar writes "hot pour"). They are the apothecary jar with dome lid and the 150 g lidded family in clear, black, red, pink, orange and amber; the amber 150 g is out of stock on 26 September 2026. Read the words: they describe pouring hot wax into the jar. They say nothing about a flame burning inside it for hours, or about the last centimetre. Many other listings say "withstand high heat" or "heat resistant" with no figure at all, and some say nothing. No CSI listing, and no data CSI publishes, covers a burn-down.

We are not saying these jars fail. We are saying nobody, including us, has published evidence that they pass the final third with your wax, your wick and your oil. How to read heat wording on any jar you find online, including a supplier's "candle-safe" claim, starts at the glass-suitability hub.

The burn-down test: taking a jar to the stop line

This is a minimum, not a full approval programme; the complete protocol for a 500-piece order is in How Do I Test a New Glass Candle Jar for Heat Safety Before Ordering 500 Pieces?, and the full programme a small brand should keep is in What Jar Safety Tests Should a Small Candle Brand Complete Before Buying and Selling a New Container Commercially?. Run it on a heat-resistant surface such as a ceramic tile or metal tray, in a room without draughts, with nothing flammable nearby, and never leave a test candle unattended. Keep children and pets out of the room.

Late-burn signs: stop, adjust or reject

A burn-down test produces observations, not verdicts. Here is how we would read the common ones. The action column is the order we would try things in: stop first, change one thing, retest.

Two cautions. First, one clean jar is a single data point; the pairing passes on three, and a single crack in any of them outweighs the other two. Second, a pass is specific: change the wick, wax, oil, load, fill weight or the jar batch and you are back to an early-pass-only candle. Mass-produced glass varies between batches, so we would repeat the burn-down on three jars from every new batch you intend to sell.

What a burn-down test needs

Five items, chosen on what their pages state, price and stock on 26 September 2026, not on test results: CSI has none for these jars. The oil for costing is British Rose Fragrance Oil (page: "up to 10% soy"; 100 g ₹342.20, ₹3.42 per gram), but use the oil you will sell: oils can change how a wick behaves, so a burn-down with a different oil proves less.

What the test costs, and a Diwali timeline

One test candle costs about ₹214.93 in materials (jar ₹94.40, 138 g of wax ₹70.02, 12 g of oil ₹41.06, a C2 wick ₹9.44), excluding labour, freight, packaging and GST. The six burned to the stop line use ₹1,278.93. Compare that with 100 of the same jars ordered untested: ₹9,440.00, or 3.9 times the whole kit. The kit is 26% of that jar order alone, before wax, oil, labels and your time.

If all three jars on a wick pass, production of 100 candles at 150 g and 8% looks like this in our worked example: 100 jars (₹9,440.00), 15 kg of CSI 464 for the 13.8 kg needed (₹7,611.00), 1.2 kg of British Rose as 1 kg plus two 100 g bottles (₹3,752.40) and 100 C2 wicks (₹944.00): ₹21,747.40, or ₹217.47 per candle, excluding labour, freight, packaging and GST.

A dated plan for Diwali, 8 November 2026

Diwali falls on 8 November 2026, 43 days after these prices were checked. A burn-down takes days, not an afternoon, because each jar needs several sessions with full cool-downs between them. Dispatch is from Pune and delivery time depends on your location, so confirm it on WhatsApp.

If deliveries run late, shorten the cure on the test jars, never the burn-down. A candle that cracks in a customer's home in the last week of its life is a worse Diwali result than a candle that shipped three days later.

Frequently asked questions