My Vanilla Candle Sweats at 10% but My Lavender Candle Does Not — Should Every Fragrance Have Its Own Maximum Practical Load?

Same ceiling, different oils: what the pages actually say

Start with the paperwork, because it explains why this question arises at 10% and not at 12%. There are two ceilings on every candle: the wax page's and the oil page's, and the lower one governs. In your pair they happen to be the same number.

Two things follow. First, nothing on either page gives vanilla more room than lavender: both stop at 10%. Your vanilla has not broken a rule; it has reached a limit the rule allows. Second, neither page tells you whether its percentage means share of total candle weight or share of wax weight. CSI's own maths uses total weight (150 g at 10% = 15 g of oil and 135 g of wax), and this page plans on that. If a ceiling were meant on wax weight, 10% of total is 11.11% of wax, just past it, which is one more reason 10% is the edge for both oils rather than a comfortable working load. The basis maths itself lives in the load-basis hub; the audit of what your own batch sheet did is the basis-audit post.

Why one oil sits happily at the edge and another doesn't is a real question with an honest short answer: fragrance oils are mixtures, their composition is not published, and a given wax holds some mixtures less readily than others. You observe that, you don't predict it. The range-wide version of that question, every scent at the same load, is the same-load range map. This page does the narrower job: turning one oil's failure into one number you can write down and pour to.

Vanilla sweats, lavender dry: what the split usually means

Before the difference is blamed on the oil, check that the oil is the only difference. The table ranks the explanations we would consider, most likely first. The ranking is editorial judgement, not measured data; CSI publishes no failure statistics. What matters is the right-hand pair of columns: the observation that rules each one in, and the one that rules it out.

Rows 2 to 6 are why the ladder has a fresh 10% control. If you pour the control in the same session as the lower rungs and it doesn't sweat, the old candles failed for a reason the new session didn't reproduce, and the ladder can't be read as a load result. Row 1 is the case this page is built for, and it is the most common in our judgement when both oils really were made side by side.

The practical-max card, field by field

A card is a single page (or spreadsheet row) per oil that answers one question: at what load is this oil poured in this wax, and why? Most of its fields are copied from product pages. Only two are yours, and only one of those needs a test.

Three rules keep the card useful. Field 5 is a measurement, field 6 is a choice. Lavender being dry at 10% fills field 5; it doesn't oblige you to pour at 10%. Whether 8% throws as well as 10% for lavender, or for vanilla, is the separate question of optimising each oil's load for throw, and the card simply records the answer. Field 5 can never exceed field 3. There is nothing to test above the governing ceiling, and no card should carry a number above it. A card belongs to one wax. Pour the same vanilla in Premium Coconut Soy Wax CSI 468 ("up to 13%") and you have a new card with a new field 2; the Vanilla Bean page's "Up to 10% for soy wax" would still cap field 3 at 10%.

The same card works for any vanilla in CSI's range, and the ceiling logic is what changes. Dark Vanilla states "candles (6-10%)" and "Vanillin present - note for light-coloured wax"; Dark Vanilla Oud states "Recommended fragrance load 6-10% for candles"; Cinnamon Vanilla states "Up to 13% load for candles", yet in CSI 464 its field 3 is still 10%, because the wax is the lower ceiling. A higher oil-page maximum never lifts a wax ceiling.

The vanilla-only sweat ladder

The ladder fills field 5 of the vanilla card. It changes only the load; wax, jar, wick, add and pour temperatures, stir, cooling spot and storage shelf all copy the batch that sweated. For the full sweat bracket across loads up to 12% where a wax allows it, see the practical-capacity bracket; this ladder is the short version for one oil already failing at the ceiling.

Add seven jars (₹660.80) and seven C1 wicks (₹41.30) and the ladder's material value is ₹1,878.03, excluding labour, freight, packaging and GST. Starting with nothing, the cash outlay is ₹2,300.40 (100 g oil, 1 kg wax, ten jars, ten wicks); if you own the wax, jars and wicks the batch used, it is the ₹790.00 bottle. Two 50 g bottles (₹820.00) would cover it too but cost ₹30.00 more than one 100 g. The fresh 10% candle is a control that reproduces the formula you already poured, not a load to lock: it sits at CSI 464's ceiling on CSI's total-weight reading and past it if the ceiling means wax weight, which is one more reason it may sweat. Your old 10% candles stay in the test as a second 10% reference; they are at least as old as the new rungs, so if anything they have had longer to sweat.

Reading the ladder and locking the load

In the worked example drawn above, the fresh control and the old candles sweat, 9% shows a film in one duplicate, and 7% and 8% stay dry in both. Field 5 reads 8%, and so does field 6. What does locking 8% instead of 10% change on the bill? Per 150 g candle, 12 g of vanilla instead of 15 g, 20% less oil. At the 1 kg price (₹7,646.00, about ₹7.65 a gram) that is ₹22.94 of oil per candle, or ₹2,293.80 over 100 candles. A full 8% candle in CSI 464, the 150 g jar and a C1 costs ₹262.07 in materials against ₹283.49 at 10%, a difference of ₹21.42 once the extra 3 g of wax is paid for; both figures exclude labour, freight, packaging and GST.

Lavender's card is simpler. Field 5 reads 10%, from candles you have. Whether field 6 should read 10% or less is partly a basis question (9% is the highest load inside both readings of CSI 464's ceiling) and mostly a throw-and-cost question: 10% lavender uses ₹81.00 per candle at the 50 g price against ₹64.80 at 8%, ₹16.20 more per candle. If a blind burn can't tell them apart, the 8-versus-9% tie-breaker shows how that argument usually ends.

What not to buy for a vanilla that sweats

And the purchase we would argue against hardest: don't ladder every oil in your range. At about ₹1,878.03 of materials per oil, laddering twenty scents is a four-figure spend per scent to learn, in most cases, that they were already fine at your house load. Cards for all; ladders for the few that fail.

What the two cards need

Five items: the oil that needs a ladder, the oil that doesn't, the wax whose ceiling both cards share, and the jar and wick held constant. Chosen on page statements, price per gram or unit, and stock on 26 September 2026; CSI has no measured throw or sweat data, so nothing here is ranked on performance. Oil tables show every size.

Frequently asked questions