I Want to Switch from 8% to 10% Fragrance Load Across My Collection — Do All My Previously Approved Wick Combinations Need to Be Retested?

What two percentage points actually change

On CSI's convention, load is a percentage of the total candle weight. A 150 g candle at 8% is 12 g of oil and 138 g of wax; at 10% it is 15 g and 135 g. The candle weighs the same and fills to the same line (oil density differs a little from wax, so confirm with a test pour). What changes is that 3 g of wax have become 3 g of oil in every candle, and the wick now draws a different fuel. If you have been working "of wax weight", the bases are explained in the load-basis post; this page stays on total weight.

What that does to a burn is a tendency, not a law, which is why it is tested rather than predicted. More oil can make a flame smaller or sootier; heavy, viscous, vanilla-rich or resinous oils are the ones most commonly linked to carbon build-up on the wick (mushrooming); lighter florals and herbals often burn more easily. A wick that was comfortable at 8% may stay comfortable, may start to mushroom, or may change how quickly the pool reaches the glass. None of CSI's oils has measured wick data, and CSI publishes no wick-performance data at all, so no page can tell you which way your candles will go.

Two things therefore follow. First, an old 8% approval is evidence about a neighbouring candle, not this one: it tells you where to start and how much risk you carry, not what to sell. Second, the size of the retest should follow the size of that risk. A collection-wide move is many small load changes at once, and treating every combination the same wastes candles on the easy ones and under-tests the hard ones.

Gate 1: the two ceilings, wax by wax and oil by oil

Every candle has two stated limits: the wax maker's maximum and the oil page's candle maximum. The lower one governs. Both are formulation limits stated by the product pages, not regulatory limits, and a maximum is not a recommendation: many candles perform best below it. Run this check on paper for every combination before you pour a single test candle.

That first row is the uncomfortable one. CSI 464 is a popular container wax, and its page's "up to 10%" means a collection moving to 10% in CSI 464 is moving to the wax's stated maximum on our own convention, and slightly past it if the maker meant wax weight. That is allowed on the page's wording, but there is no margin left. A 1 g slip on a 15 g target puts a 150 g candle at 10.67%, past the ceiling on any reading; at 9% the same slip lands at 9.67%, still inside it. We sell the extra oil a 10% collection needs, and we would still suggest 9% for CSI 464 collections, unless your weighing is good to the gram every time. Moving to a wax with a higher stated ceiling does not skip the retest either: a wax change restarts the wick evidence on its own, which is the soy-to-soy-coconut carry-over post.

The oil side of the check

At least 48 of CSI's 102 oil pages state a candle load or range. Of the 48 we captured in structured form, 42 top out at 10%, three at 9%, one at 11% (Autumn) and two at 13% (Japanese Cherry Blossom and Cinnamon Vanilla). Several more mention a candle figure only in FAQ or usage text, so read your own oil's page rather than our tally.

Gate 2: blocked, short screen or full life

Once Gate 1 has removed what cannot move, sort every remaining approved combination by how much it is likely to change and how much it matters if it does. The rules below are our judgement, built from the tendencies above. Write your own version down before you pour, and do not soften it after the results come in.

Why a control at 8%? Because your 8% approval was poured months ago, from a different wax lot, cured on a different calendar and burned in different weather. Comparing a fresh 10% candle with an old notebook confuses the load change with everything else that changed. One 8% candle poured the same day from the same melted wax makes the comparison fair. Why one neighbouring rung, and which? Go in the direction the 8% record leaned: a fast pool or a flame at the top of its range points down a rung; a slow edge points up. If the record had no lean, we test the rung down, our judgement, because a late-burn heat failure costs more than a slow first burn. CSI's rungs are C1 → C2 → Cotton Braided, with no sized, tabbed wick between C1 and C2; the full grid logic is in the behaviour-groups post.

The same sorting idea, applied to a jar change rather than a load change, is the 7.5 to 8 cm retest-scope post. The difference here is that Gate 1 can block a combination outright, and that load affects every jar in the collection at once.

Running the screen and the full-life track

Two specific situations already have their own posts in this series. A rose that pools very deep at 10% and the choice between stepping the load down and stepping the wick down is the 10% rose route comparison. A coffee that mushrooms at 10% and what a lower load costs in scent is the coffee load step-down. If an unscented approval is all you ever had, start with the unscented-to-scented re-bracket instead of this plan.

Nine combinations in candles, grams and rupees

Our worked example is a collection of nine approved combinations on CSI 464, all at 8%: eight oils in the 150 g Clear Glass Jar and Lavender also in the 500 g Apothecary Glass Candle Jar with two wicks. Fills are 150 g and 500 g (the Apothecary page states "Fill capacity: ~500 g - 700 g (2 sizes)"; we use the smaller). Records, wicks and leans are illustrative.

Dark Vanilla's rung up is Cotton Braided, whose own page describes it as "engineered for large-diameter candles exceeding 8 cm", which a 7 cm jar is not. That is why it is a separate, labelled arm rather than a routine step, and why a fail on both sides there would send Dark Vanilla back to 8%. Grams per candle: 150 g at 10% = 15 g + 135 g; control at 8% = 12 g + 138 g; Apothecary at 10% = 50 g + 450 g, control 40 g + 460 g.

Most brands making this move already stock the oils and jars, so the cash that leaves the bank is usually the wax and the wick packs: ₹3,731.20. Against the baseline of putting every legal combination on the full-life track (45 candles), the plan pours 11 fewer candles; more usefully, it carries 23 candles to full life instead of 36, and full-life sessions are where the weeks go.

What 10% costs per candle, oil by oil

Going from 8% to 10% in a 150 g candle adds 3 g of oil and removes 3 g of wax. At production prices (each oil's 1 kg, CSI 464 at ₹0.51 a gram) the wax saving is ₹1.52 a candle, so the net change is the extra oil minus that.

Three things in that table matter more than any single figure. The dearer the oil, the dearer the move: Raat Ki Rani and Dark Vanilla add two to three times what British Rose or Citrus Lemon do. The rupees scale with volume: for Raat Ki Rani, 100 candles take ₹10,161.00 of oil at 10% against ₹8,128.80 at 8%. And your oil purchasing changes shape: 1 kg makes 83 candles at 8% and 66 at 10%, so you buy 25% more oil for the same candle count. A 9% step on Raat Ki Rani would add ₹9.40 net per candle (13.5 g = ₹91.45).

The large jar scales the same way. Lavender in the 500 g Apothecary goes from 40 g to 50 g of oil, ₹47.14 more per candle net of the wax it replaces. Large, multi-wick candles are where a load change costs most per unit and where a late-burn failure is most expensive, one more reason they sit on the full-life track.

When 10% is the wrong move

A collection-wide load increase usually starts with customers saying "make it stronger". Before committing, be clear about what 10% can and cannot do. More oil does not raise throw in a straight line, and past what the wax holds it tends to cause sweating and can hurt the burn. Many makers find a candle that was weak at 8% is only slightly stronger at 10%, sometimes not noticeably so; the 8% to 10% no-improvement post works through why. If the real complaint is room size, a ceiling fan or a too-small jar, a load change will not fix it; the stronger-candles roadmap ranks the other levers by cost.

A partial move is also a legitimate end state. If four of your nine combinations pass at 10% and five do not, you now know which candles respond to more oil and which do not, and you can run the collection at two house loads with the reasons written down. The purchase gate that should come before any 1 kg or 1,000-wick order is the 1,000-wick purchase gate; the full-life calendar for a 150 g, 7 cm candle is the full-life calendar post.

What to buy, and what to skip

Five items, chosen on page statements, price and stock on 29 September 2026. CSI publishes no wick-performance data, so nothing is ranked on burn results. Candle counts in the tables use 12 g (8%) and 15 g (10%) of oil per 150 g candle.

Frequently asked questions