My Candle Has Been Curing for Two Weeks but Still Smells Weak When Burning — Should I Change Fragrance Oil, Wax, Wick or Fragrance Load First?

What two weeks of cure has already ruled out

Soy commonly needs a cure period before hot throw is judged, and oil pages that state one say 48–72 hours; the Roasted Coffee page says "Always cure soy wax candles for at least 48-72 hours before burn testing". Many makers judge at day 7 and again at day 14 as common practice. By day 14, waiting longer is unlikely to rescue a candle that is clearly weak (whether longer cure keeps helping is a plateau question for the formulation series), so the next move is a change. The question is which, and that is what this page owns. When to stop waiting is the day-14 decision tree; this page takes over once you have decided to change something and need to choose what.

Before any lever, two free checks, because a ladder built on a bad test climbs in the wrong direction. First, was the day-14 burn judged fairly: a closed room, fan and AC off, a fresh nose walking in once the pool was full? If not, re-judge before changing anything; the cluster hub walks through the evidence you can read from burns already done, and the fade-point test covers a candle that fades within a burn. Second, was the load what you think it is: weighed in grams on total weight (150 g at 8% = 12 g of oil + 138 g of wax), added at the wax page's temperature and stirred thoroughly? A capful or "80 g per kg of wax" is a different load; the load-basis hub converts them.

The change-order ladder, costed

The ladder prices each lever as a round of two 150 g test candles (a duplicate, so one odd candle cannot decide), burned against a retained twin of your current candle as the control. The worked example is a Roasted Coffee candle at 8% of total weight in CSI 464, in the 150 g Clear Glass Jar with Golden Lid with an Eco C1 wick. Swap in your own materials; the structure holds.

Read the third column first: the dearest round costs ₹98.04 more than the cheapest. A pour is a pour; two jars, 276 g or so of wax and about 24 g of oil cost roughly the same whatever you change. So "cheapest" on this ladder is really about the other three columns: the cash you must spend before you can test, the number of things you must check again if the change works, and what it does to every candle you make afterwards. By all three, the wick step is the lightest lever and the oil swap the heaviest.

Where the round costs come from

Rung 1: ₹110.40 of Roasted Coffee (24 g at the 50 g price) + ₹140.04 of CSI 464 (276 g) + ₹188.80 for two jars + ₹18.88 for two C2 wicks = ₹458.12. Rung 2 at 9%: ₹124.20 of oil (27 g) + ₹138.52 of wax (273 g) + ₹188.80 + ₹11.80 = ₹463.32. Rung 3: ₹110.40 + ₹179.40 of CSI 468 + ₹188.80 + ₹11.80 = ₹490.40. Rung 4: ₹215.52 of Coffee Vanilla Tobacco at its 50 g price + ₹140.04 + ₹188.80 + ₹11.80 = ₹556.16. Across a production run the last column matters more: over 100 candles the load step adds ₹560.70 of oil at the 1 kg price, the wax swap ₹1,788.48 at 10 kg prices, and the oil swap ₹5,833.20 at 1 kg prices.

Which rung your evidence supports

The rule is "cheapest, least disruptive lever the evidence supports". A rung the evidence rules out is skipped, however cheap it is. The table maps the four lines from your day-14 burn to the first rung worth running. The pool times and depth are commonly cited rules of thumb, not CSI data; for the 7 cm-wide 150 g jar, the rough "one hour per 2.5 cm of width" gives about 2.8 hours.

The ceiling matters for rung 2. On CSI 464 the wax page states up to 10% and Roasted Coffee's page states 6–10%, so the lower of the two is 10%. The wax page does not say which basis its 10% uses; read conservatively on wax weight it allows about 9.09% of total weight, and 9% of total weight is 9.89% of wax weight, which satisfies both readings. That is why the load step here is 8% to 9%, not to 10% (11.11% of wax weight). Never go above the lower stated maximum. The tie-breaker between 8% and 9% when sweating is also in play is the fine-at-8%, sweats-at-10% post; the cost of each 1% before a bulk order is the dosage-before-the-kg-order post.

Running a rung: two candles, one change

For a scale-up version of the same idea, testing several levers for maximum throw without adding oil, see the lever-by-lever plan. The first-level version of this page, ranked by cost and effort, is wax, wick or oil first.

The case against buying a new oil first

A weak candle makes an oil feel guilty, and a new bottle feels like progress. Here is why, in our judgement, it is the wrong first purchase for most two-week-weak candles, argued against our own interest.

The cost of getting the order wrong is concrete. Buy the oil swap first, watch it fail in a partial pool, then run the wick step you needed anyway, and you have spent ₹1,014.28 of materials on two rounds and 28 days of cure on a question one round could have answered. Run the wick step first and, if it fixes the pool, the original oil often turns out to be fine; if it does not, you climb with the pool question settled.

If the candle has several problems at once, tunnelling, soot and a weak throw, do not climb this ladder yet: the order of fixing multiple faults is the five-problems post, because an upstream fault can cause the downstream ones.

What each rung needs from CSI

Five items in ladder order: the wick for rung 1, the wax for rung 3, the worked-example oil you already own for rung 2, the wick rung above C2, and a worked-example oil for rung 4. Chosen on page statements, price per unit and stock on 26 September 2026; CSI has no measured throw data, so nothing is ranked on throw. Oil tables show every size, 150 g candles at 12 g and two-candle rounds at 24 g.

Frequently asked questions