You have three "separate" candle problems. A black ring at the rim. A river of wax running down the glass. A jar split on the second burn. You will see them diagnosed as air flow, wick length, and bad glass, in that order. Wrong, wrong, and half-right.

They are one problem wearing three hats: the wick diameter does not match the jar diameter and the wax. The wick is the only part of the candle that is sized in millimetres and chosen in vibes, and it is the single variable that decides fuel flow, pool diameter, soot, and heat load. Everything you are blaming the room for is the cord talking.

The wick is a fuel pipe, not a match

A wick moves liquid wax by capillary action, and its bore is a throttle. Small bore, thin stream, small flame, small pool. Big bore, thick stream, roaring flame, wide pool, and a burn rate that will finish the wax before you are done with the room. That is the entire physics. The flame is a combustion engine with one control knob, and the knob is wick diameter.

Here is what nobody puts on the label: a single wick code does not work across jar widths or wax types. CD6, CD8, CD10, CD12 are cotton wick codes where the number is the plied/size index; wood wicks run 6/10/12 mm. The common failure is the "one wick fits all" batch from the factory — same CD8 in a 60 mm jar, a 90 mm jar, and a 120 mm tumbler, in soy, paraffin, and palm. One of those three is correct, probably none of them, and the other two will fail in predictable directions: the narrow jar smokes and mushrooms, the wide jar tunnels and starves.

The test that replaces every other test: burn two hours, measure the pool

Forget the burn time, the scent throw, the "premium" claim. Light the candle, let it run an unbroken two hours, and look at the pool. Done right, the melt pool reaches the glass wall all the way around before the two-hour mark. That is the only pass/fail line that matters, and it is the same one the factory uses on the test bench.

Two failure signatures, read them in the glass:

Narrow well, deep in the middle, dry edges after two hours. Wick too small for the jar. The pool never reaches the wall, so the edge never melts, and on the next burn the cold wax rim insulates the center into a tunnel. This is the tunnel, and it is not a wax problem, it is a bore problem. Fix: step the wick size up one code, or for a soy in a 90+ mm jar, go to a wood wick at 10-12 mm, which throws a wider, hotter pool.

Pool at the wall within 30 minutes, heavy soot, wick mushrooming black, maybe a hot-spot crack in thick glass. Wick too big. The flame is outpacing the wax it can feed cleanly, so it runs rich: soot, smoke, and a heat load that a 5+ mm glass wall can store and then release as a stress fracture. Fix: step the wick size down one code, and trim to 5-6 mm after every burn so the excess wick stops acting as a second, uncontrolled flame.

If you are not sure which side you are on, the soot ring tells you: black rim = too big, tunnel = too small. Both can happen in the same box, because the factory shipped one wick code and three jar sizes.

Why "smokeless" and "clean-burning" wicks sell so well

Because the buyer is treating a symptom. A candle that smokes is being sold a fix, when it is being sold a misconfiguration. The borax or salt soaking that "treats" cotton wicks is real — it stiffens the braid, steadies the flame, and does cut soot a bit — but it changes flame character, not fuel flow. A CD8 that is wrong for a 120 mm jar does not become right because it was dipped in borax water. It burns steadier and wrong, which is somehow worse, because it stops announcing itself.

Same story with wick length as a substitute for wick size. Trimming is a maintenance act for a correctly sized wick. If you have to trim to 4 mm to stop the smoke, you do not have a trimming problem, you have a bore problem. The trim buys you a clean burn and a shorter candle life, and the two are not the same thing.

The failure modes, mapped to the one knob

Black soot ring at the rim. Wick too big, or untrimmed. The flame starves for oxygen at the tip and runs rich. Down one wick code, trim to 5-6 mm before every burn, and make sure the jar is not sitting in a draft that is pinching the air.

Tunnel: a river down the glass, dry rim. Wick too small for the diameter. The pool never reaches the wall. Up one code, or wood wick in wide soy vessels. This is the one buyers mistake for "the wax is low quality," and the wax is usually fine. It is being starved by its own wick.

Crack in a thick glass jar, first or second burn. Heat load out of proportion to the glass. A wick that is one or two codes too big puts the pool at the wall in 20-30 minutes and holds it there; the 5+ mm glass wall takes hours to cool, and the temperature delta across that wall is the fracture. The wick was wrong. The glass was honest about it.

Mushroomed, blackened wick tip. Uncontrolled soot from an oversized wick or a too-long wick. The fix is the same as the soot ring: smaller bore, shorter wick. If you keep it, trim every single burn. No exceptions.

Wax runs down the outside. Pool out of control, wick too big, and usually a thin glass wall that cannot store the heat without passing it to the surface. Down one code. If the glass is under 3 mm, the honest fix is a different vessel, not a different wick.

The sizing logic, in the order that matters

One, measure the interior diameter of the finished vessel, not the outer. Two, pick the wick family for the wax — cotton for most container soy and paraffin, wood for wide-diameter soy and palm where pool width is the whole job. Three, start from the middle code of that family and run the two-hour pool test. Four, adjust one code at a time, in the direction the pool tells you, and do not move to the next burn until the pool is hitting the wall by the two-hour mark. Five, if the pool is right but the glass is thick, you are done; if the pool is right and the glass is thin, you are about to learn what a hairline crack is, and you should have picked a different wick or a different jar.

That is the whole job. Two numbers — jar diameter, wax type — and a wick code. Everything else is marketing, or is a symptom of getting those three wrong.

If you cannot be at the candle, use a candle that does not burn

Hospitality, front desk, a lobby you are not standing in: the wick-sizing job becomes a liability you cannot supervise. That is the whole argument for the flameless side of the counter — a LED candle range that holds a 2700K pool without a single millimetre of wick to get wrong. You lose the smell, which you were never going to supervise anyway, and you keep the look that sells the room. For the outdoor run where the flame was never the point, a warm-white light set rated for the weather does the same mood work with zero melt pool to manage.

FAQ

How do I know if my wick is the wrong size? Burn two hours and watch the pool. If it never reaches the glass wall all the way around, the wick is too small for the jar. If the pool is at the wall within 30 minutes and the rim is sooting black, the wick is too big. The pool is the instrument, not the flame.

Why does my candle tunnel down the middle? Wrong wick, too small for the diameter. The pool is not reaching the wall, so the rim stays solid and insulates the center on the next burn. Step the wick up one code or go to a wood wick in wide soy vessels, and let the first burn run until the pool reaches the edge.

Can I re-wick a finished candle? You can, and the results are better than the candle you are disappointed in. Pull the wick, clean the base, fit a new wick one code in the direction your failures pointed, and you will burn the same wax correctly. The wax was not the problem. The cord was.

What wick should I use for soy in a 90+ mm jar? Wood wick in the 10-12 mm range, or a cotton one step up from the middle code. Soy in wide vessels starves narrow cotton wicks, and that is where most "my soy smells weak and tunnels" complaints are actually coming from.

Does wick size change burn time? Directly. Bigger bore, more fuel flow, faster burn. That is why a "premium" candle in a thick glass jar can finish in hours — the wick was sized to the flame, not to the jar, and it is eating the wax twice as fast as it should be.