You've followed every other rule. You poured at the right temperature. You used a soy blend that holds 9% oil. You let it cure for 14 days. And then the first burn happens and the candle either drowns in a blackened, sooty mess or the melt pool stalls out before it reaches the glass wall, stranding half the wax against the side of the jar.

Every time this happens in a production run, the first instinct is to blame the wax or the fragrance oil. Those are the variables you can see. The wick is the variable nobody looks at.

Here's the thing about the wick: it's not a generic component. It's a sizing decision with a hard numeric spec, and most candle shops — and most of the small-batch operations I've audited — have never checked that their wick's rated diameter range actually matches the container they're pouring into.

The Spec You Never Wrote Down

Every wick manufacturer publishes a sizing chart. It's usually a simple table: wick size → container ID (inner diameter) range → recommended wax type. A CD8 is rated for roughly 28–40 mm containers in soy. A CD10 covers roughly 38–50 mm. An Eco-10 handles 40–54 mm. These aren't suggestions. They're the thermal capacity limits of the wick.

The ID is the inner diameter of the container, not the outer. And it matters more than you'd think because the glass thickness of a 3-inch jar versus a 3.5-inch jar can shift the ID by 2–4 mm, which is enough to push a CD8 into the "undersized" zone for that specific vessel.

Here's the production failure I see most often: a shop sources a wick that works perfectly in their 8-ounce amber jar, then switches to a 10-ounce tumbler with 2 mm more ID, and calls the same wick "fine." It's not fine. The wick is now thermally undersized for the new container. The melt pool forms, but the heat output can't drive it to the glass wall within the burn window. You get a partial melt pool, a tunnel forms in the next burn, and by burn three the candle is a wax ring with a dead center.

What "Undersized" Actually Looks Like in a Jar

Undersized wick, correct wax, correct fragrance load. The first burn looks normal for the first 20 minutes. Then the melt pool stops expanding. You can see it: a circular pool of liquid wax sitting in the center of the jar, with a visible solid ring of un-melted wax between the pool edge and the glass wall. The wick is burning steady, no soot, no smoking. It's doing exactly what an undersized wick does: producing less heat than the container geometry demands.

The fix people try first is "burn it longer." That's the wrong fix. Burning longer with an undersized wick doesn't fix the thermal deficit. It just extends the time the candle takes to reach a full melt pool, and in the meantime the fragrance is volatilizing from a smaller surface area. You lose throw. You lose burn time. And the second burn starts with a deeper, narrower pool that tunnels even harder.

What "Oversized" Looks Like — and Why It's Worse

Oversized wick, same container. The first burn is dramatic. The melt pool reaches the glass wall in 45 minutes instead of 90. The flame is tall, bright, and — if you're paying attention — the glass is getting uncomfortably hot at the top rim. The wax is being consumed faster than the wick can capillary-feed it, so the flame is running on a thin, volatile-rich fuel stream. That's the tall, flickery flame you see in a poorly sized candle. That's also why oversized wicks soot. Not because of the wax. Because the combustion zone is too hot and too oxygen-starved at the wick surface.

In a 3-inch glass container, an oversized wick can push the glass temperature past the point where the adhesive bond holding the wick tab to the base begins to soften. You pull the candle out of the jar and the tab is on the glass, not the wick. That's a product failure that shows up in returns and 1-star reviews as "the wick came loose."

The Sizing Check That Takes 90 Seconds

Before you commit to a wick for any new container, do this:

1. Measure the ID. Calipers, not a ruler. Measure the inside diameter at the top of the container, where the melt pool will form. Write it down. If you're switching glass suppliers, re-measure. "The same 8-ounce jar" from two different glass vendors can have 1.5 mm of ID variance.

2. Cross-reference the wick chart. Find the wick size whose rated range includes your measured ID, with at least 1 mm of margin on both sides. If your ID is 38 mm, you want a wick rated for 36–42 mm, not one rated 34–38 mm where you're sitting at the absolute top of the range. Being at the edge of the range is where you get marginal performance that looks "mostly fine" in the test burn and falls apart in production at scale.

3. Do the 2-hour test burn in the actual container. Not a sample. The actual jar. Measure the melt pool diameter at 60 and 120 minutes. At 120 minutes, the pool should be within 3–5 mm of the glass wall. If it's 10 mm or more from the wall, the wick is undersized. If the glass is hot to the touch at the top rim at 90 minutes, the wick is oversized.

4. Check the tab bond. After the test burn, pull the wick tab out of the wax. If the tab stays on the glass, your adhesive is the problem, not the wick — but you need to know which one failed before you try to fix the wrong thing.

The Wicks That Don't Fit the Chart

There's a class of wicks that the standard sizing charts underspecify: paper-wrapped cotton wicks (the CD and ECO series) in containers with thick glass walls or high thermal mass. The glass absorbs heat from the melt pool that a thinner-walled container wouldn't. In practice, a wick that's "correct" for a 3 mm glass jar may need to go up one size for a 5 mm glass tumbler, even if the ID is identical. The chart doesn't account for thermal mass. You have to account for it.

Same issue with dark-colored glass. Amber and teal glass absorbs and re-radiates more heat than clear glass of the same thickness. If you're running amber 8-ounce jars and clear 8-ounce jars side by side with the same wick, the amber ones will run slightly cooler at the melt pool surface and the clear ones will run slightly hotter. One wick size is "correct" for one of those jars. It's not correct for both.

What to Do When the Wicks You Have Don't Match

If you're stuck with a wick that's one size too small for a container you've already poured 200 jars of, you have two options and neither is great.

Option 1: Reduce the fragrance load. A wick that's undersized is already producing less heat than the geometry demands. Adding more fragrance oil to compensate makes it worse — the oil raises the wax's pour point and changes the melt behavior in a way that makes the undersized wick even less capable of driving a full pool. This is the "add more oil to fix the throw" trap, and it's making the wick problem worse while hiding it.

Option 2: Accept a shorter first burn and manage customer expectations. If the wick is only marginally undersized — the melt pool reaches the wall in 2.5 hours instead of 1.5 — you can ship it, but your burn time claim on the label needs to reflect the actual performance, not the theoretical maximum. A 10-ounce candle that actually burns 28 hours instead of 40 is a different product. Selling it as 40 is how you end up with a return rate that kills the SKU.

Neither option is a real fix. The real fix is to source the correct wick size before the next pour. That costs you one test burn and one afternoon. The alternative costs you a batch.

FAQ

How do I know if my wick is the wrong size?If the melt pool doesn't reach within 5 mm of the glass wall after 2 hours of steady burning, the wick is undersized. If the glass is hot to the touch at the top rim before the 2-hour mark, it's oversized. The melt pool measurement is the only reliable indicator. Burn time alone tells you nothing until you've measured the pool.

Can I use the same wick for different container sizes?Only if the container IDs fall within the same wick size's rated range. "About the same size" isn't a spec. Measure the ID. Cross-reference the chart. The 2 mm that separates a CD8 from a CD10 is the difference between a working candle and a tunneling one.

Why does my candle soot even though the wax is natural?Soot is a wick sizing and airflow problem, not a wax purity problem. An oversized wick produces a flame that's too tall and too hot, which causes incomplete combustion at the wick surface. You can have 100% soy wax and 100% essential oils and still soot heavily if the wick is one size too big for the container.

What's the right wick for an 8-ounce soy candle?It depends on the ID of your specific 8-ounce jar, the glass thickness, and the glass color. As a starting point, a CD8 or ECO-8 in a 30–34 mm ID clear glass jar is a common match. But "8 ounces" is a volume, not a diameter. Two 8-ounce jars from different suppliers can have 2 mm of ID difference. Measure the jar, not the label.

Does wick size affect fragrance throw?Indirectly, yes. An undersized wick produces a smaller melt pool, which means less surface area for the fragrance to volatilize. You don't smell the candle because the wax isn't melting wide enough. People assume it's a fragrance load problem and add more oil, which makes the melt behavior worse without fixing the pool size.