My New Candle Jars Are Only a Few Millimetres Wider than the Old Batch — Can That Difference Be Enough to Change Melt-Pool Performance?

The arithmetic: 2, 3 and 5 mm on a 7 cm jar

Melt-pool area is π × r², using the internal radius. The table works the numbers for a 7.0 cm internal diameter, a worked example (the White Matte page states a "width 7 cm", which is not said to be internal; if it is outer, the inside is smaller, so 7.0 cm is a round worked figure, not this jar's). Two things stand out. The area changes about twice as fast as the diameter: 2 mm is 2.9% wider but 5.8% more area, and 5 mm is 7.1% wider but 14.8% more. And the same arithmetic runs the other way. A jar 2 mm narrower has 5.6% less area, so the same wick may now run a deeper, hotter pool.

What the arithmetic cannot tell you is whether your wick had room to spare. A wick that reached the edge of the old jar comfortably, with a calm flame and a pool about 0.5–1 cm deep (a commonly cited rule of thumb), may shrug off 2 mm. One that only just made it will not. That is a burn result, not a calculation. The single-difference explainer, what 5 mm does and when to retest, is the 5 mm post; this page is about a delivery that varies, and how to test it at its edges.

What a few millimetres looks like in the burn

The table ranks what we would suspect when candles in a new, slightly wider delivery burn differently, most common first. The ranking is editorial judgement; CSI has no jar-failure data. The right-hand columns are observations that decide it.

Measure the whole delivery: the spread sample

Where the old jar lands in the spread tells you most of the story before you light anything. Old jar near the middle: the delivery varies around your old size; "a few millimetres wider" was true of the unit you happened to pick up. The extremes still need burning, but the likely answer is a pass. Old jar at the narrow end: the whole delivery has moved wider, most often through thinner glass. Expect the widest units to be the ones that fail, and read 3268 if the supplier or the design changed. Old jar outside the spread altogether: this is a different jar in practice, whatever the listing says. Treat it as a new container, with a proper bracket rather than a spread check.

Burn the extremes: narrowest against widest

Pour the four chosen units on the same day from one melt, on your current wick, centred with a holder. Everything else stays as you make it. The worked example: White Matte Glass Jars filled to 175 g, CSI 464 at 8% of total weight (14 g of oil + 161 g of wax per candle), current wick C2, oil valued at Citrus Lemon's 100 g price (its page states "candles (6-10%)"). Stay within the lower of the oil page's range and the wax ceiling; CSI 464 states up to 10% (17.5 g per 175 g; because the ceiling's basis isn't stated, 9%, 15.75 g, is the cautious top rung). The White Matte page gives no numeric hot-pour temperature, only that it can "withstand high heat", so pour within the CSI 464 page's 80–90°C and, as a precaution (judgement), pre-warm clean jars and avoid the top of that range.

Costs are worked examples and exclude labour, freight, packaging and GST. The jars and wicks come from stock you already hold, so the new cash is usually nothing. For scale: one production candle in the worked example is ₹201.58 in materials, a 200-candle run ₹40,315.88; the test is 2.0% of that. Protocol: burn all four together in a still room, well apart, never unattended, in sessions long enough for a full pool. At 1, 2, 3 and 4 hours note the pool's diameter from above (or "edge"), and at the end of each session its depth with a skewer (about 0.5–1 cm is a common target) and the flame. Continue burn after burn to about 1 cm of wax left (common practice). The final third is where the narrowest units show whether they run too hot. For when a burn-down is deep enough to trust, see the burn-depth post.

Reading the extremes: keep, step or sort

Sorting by width is a legitimate production step, not a failure. In the worked example the two widest bins hold 4 of the 20 jars; if only those fail, a separate wick for them, or a different product in them, may be simpler than re-wicking the whole range. What you should not do is send the widest units out on a wick that left a ring in your test. A customer's first burn is where tunnelling starts, and the customers-report-tunnelling post shows how hard that is to reverse. If your spread is wide enough that you would rather have fewer sizes, the fewer-wick-sizes post makes the case for standardising on a diameter band.

Jar, wicks and centring

Four items: the White Matte jar named in the brief, the Eco wick range for the current wick and its steps, a centring device that takes wick position out of the comparison, and the Cotton Braided wick, listed so you know when not to use it. They are chosen on page statements, price per unit and stock on 26 September 2026. No CSI listing states an internal diameter, so nothing is ranked on fit.

Frequently asked questions