Everyone hands you a wick, calls the job done, and the candle tunnels or blackens the glass by the third burn. The wick is not a part number you pick off a list. It is a ratio, and it is the ratio that fails in your jar. Get the ratio right and the rest of your recipe mostly forgives you. Get it wrong and no fragrance load will save you.

The Diameter Is the Whole Job

The single number that decides wick size is the inside diameter of the container, measured across the widest point of the melt pool, not the label size. At a standard 8% soy load, a 2-inch round jar wants a CD-10 to CD-12 cotton wick. Push to 3 inches and you jump to a CD-18 to CD-20. That jump is not a typo. Between 2 and 3 inches the wax volume grows faster than the flame can carry heat, so one step in wick series buys you the difference between a pool that reaches the wall and one that starves in the middle.

Here is where people overreach. Past 4 inches, stop hunting for one bigger wick. A single CD-26 will run hot, throw soot, and still leave the corners of the pool half-melted. The working fix is to split the duty: two or three smaller wicks, evenly spaced, each feeding a share of the wax. We have watched this fail the other way in production, one giant wick in a 5-inch vessel, flame tall enough to smoke the glass, wax gone in half the expected time. Multiple wicks burn flatter and longer because the heat load is shared.

Three Multipliers Shift Your Size

Diameter gets you to the neighborhood. Three factors move you a size up or down, and ignoring them is why a soy chart misleads the moment you pour something else. First, wax type: paraffin burns hotter and pools faster, so you size down; a paraffin pour runs about 10% smaller than the same soy recipe. Beeswax melts high and needs more heat, so you size up. Coconut and para-soy sit just under soy. If you are reading a chart that only lists soy, every other wax on that page is a starting guess, not an answer.

Second, shape. A round container is the baseline. A square jar holds more wax in the corners than its diameter suggests, so it takes a size up; a rectangle takes a size up again. The corners are exactly where a flat wick starves, and that is the quiet source of most one-sided tunneling complaints. Third, fragrance load. Every point of fragrance above 6% nudges the effective diameter up by roughly 2%, because the oils thin the melt and slow the pool. A heavy 10% load in a tight jar asks for a bigger wick than the same jar at 6%.

How You Know Which Way You Missed

You do not need a thermometer to grade a wick. Light the candle and watch the first two to three hours. A correctly wicked candle reaches the glass wall in two to four hours and holds a flame under three quarters of an inch. Under-wicked, you see a deep channel down the middle, wax clinging to the walls, and a scent throw that barely leaves the room. Over-wicked, the flame climbs past an inch, soot rings the rim, and the container runs hot to the touch. The candle tells you the direction; the fix is one step toward the middle.

Two traps the charts cannot see will pull you off that middle. Hardeners like Vybar stiffen the wax and slow the pool, so they usually cost you a size up. Heavy dye loads choke the braid and do the same. Test the size your chart gave you, and test one size above it, then keep whichever builds the edge-to-edge pool in the two-to-four hour window. If the flame is fighting you outdoors anyway, a flameless LED candle removes the wick question entirely, and the same wind logic is why we keep pushing people toward low, anchored Christmas lights instead of a tall flame they lose by midnight.