My Unscented Soy Candle Burns Perfectly but Becomes Under-Wicked After I Add 10% Fragrance — What Should I Test Next?

An unscented pass is a baseline, not an approval

An unscented burn is a useful first look at a jar and a wax: whether the glass sits square, whether the wax pulls away, roughly what a wick does in that diameter. That is all it is. Once a tenth of the candle is fragrance oil, the fuel the wick draws is a different liquid, and the melt pool, flame and carbon on the wick can all change. Commonly the flame gets smaller or sootier as the load rises; sometimes it does the opposite. Why it happens, in tendencies rather than certainties, is the 7 cm unscented-to-10% post; this page is about what to buy and pour next.

So the purchasing rule is blunt: the unscented result cannot approve anything and cannot rule anything out. It does not prove C1 is the right family, and it does not prove C2 will be too big. It gives you a starting rung and a direction. The direction is up, because the scented candle was under-wicked; the starting rung is the one that failed. If your scented failure looks more like a shrinking flame than a tunnel, the flame-height ladder in the 6% to 10% flame post is the closer read before you order.

The ceiling check: why the top rung is 9%

Two ceilings apply to every candle: the wax maker's stated maximum and the oil page's stated candle maximum. The lower one governs. In the worked example both say 10%: CSI 464's page states "Fragrance Load Up to 10%", and Jasmine's states "Recommended load up to 10% soy". So far, 10% looks legal. The catch is the basis. CSI's convention, used on this page, is load as a percentage of the total candle weight; but CSI 464's page does not say which weight its 10% refers to, and the conservative reading is the wax weight.

Ceilings are formulation limits, not regulatory ones, and the maximum a wax will hold is not the load that performs best; more oil past what the wax holds tends to sweat or pool and does not raise throw in a straight line. That is a second reason the 10% candle may have under-performed. A 9% top rung removes the ambiguity: whatever the page meant, 9% is inside it. If you also want to know whether less oil throws almost as well, that is the cost-benefit in the 6% to 9% load-rise post; and if you are stepping down from 10% to 8% for good, the confirm order is the 10% to 8% retest post.

If 10% is the product: change wax, and start again

Some makers have promised a 10% candle. The CSI wax with stated headroom above that is Premium Coconut Soy Wax CSI 468: "Fragrance Load Up to 13%", "Pour Temperature 58°C – 60°C". At 10% of total weight it is at 11.11% of wax weight, inside 13% on either reading. Luxury Soy Wax Chunks also states "up to 13% fragrance load capacity". But a new wax is a new candle: the unscented C1 data from CSI 464 does not travel, nor does anything you learn in the 9% grid. We call this Path B and cost it separately below. Most readers should not take it.

Oil rupees against wick rupees

Price the three loads and the three wicks side by side. Oil is Jasmine at the 100 g price (₹4.30 a gram) and, for production, the 1 kg price (₹3.92); wax is CSI 464 at ₹0.51. Oil replaces wax gram for gram, so each extra point of load also removes 1.5 g of wax; the net figure counts that.

Put the two together and the likely outcome pays for itself. The failed candle (10% on C1) costs ₹138.90 in materials; 9% on C2 costs ₹136.75, ₹2.15 less. The wick in that candle is 6.9% of its materials. The oil decides the economics; the wick decides whether the candle works. How much the oil step matters depends on the oil: for a dear one such as Vanilla Bean a point costs ₹11.09 net, for a cheap one such as Eucalyptus Spearmint ₹4.55. Even at the cheap end, a point of oil still costs more than the C1 → C2 step.

The re-bracket order: twelve candles, four cells

The worked example is CSI's Clear Glass Jar with Golden Lid (150 gram), whose page gives an outer diameter of about 7 cm and "Safe for hot pour up to 85°C". On our labelled 0.35 cm wall that is about 6.3 cm inside, a pool of 31.17 cm²; a very rough rule of thumb (an hour per 2.5 cm) puts a full first pool at about 2.5 hours. If your jar is different, the cells stay the same and only the grams change; how to measure the inside is the external-versus-internal diameter post.

In materials the grid consumes ₹1,610.37: ₹677.25 of Jasmine, ₹833.40 of wax and ₹99.72 of wicks. Per candle: C1 at 8% ₹127.52, C1 at 9% ₹133.21, C2 at 9% ₹136.75, Cotton Braided at 9% ₹139.31.

Path B, only if the product must stay at 10%

Six more candles in CSI 468 at 10%: C2 × 3 and Cotton Braided × 3 (90 g Jasmine, 810 g wax), poured at the wax page's 58–60°C. It needs a 1 kg bag (₹650.00, 190 g left over), one more 100 g of Jasmine (three bottles leave 52.5 g after both paths) and the C2 20-pack instead of the 10, which holds both paths' C2 plus a re-pour in each wax. That adds ₹1,174.40 to the order and ₹977.82 of materials consumed. A C2 candle at 10% in CSI 468 costs ₹161.69 against ₹136.75 at 9% in CSI 464.

Reading the grid, and what each result buys

Two gates, as set out in the cluster hub's approval rule: a screen at burn 3 and a full-life pass on all three candles of one cell. Only the second releases production money. Read the cheap cells first: if the old wick works at a lower load, you need no new wick SKU at all.

Why the full jar, not a good burn 3: heat commonly builds as a candle burns down, so the wick that looks right near the top can run hot near the base. How far a test candle has to go before it counts is the acceptance-depth post. A bracket only closes when the winner has a documented fail on each side, which is why Cotton Braided is in a 7 cm grid at all: the bracket-closure rule explains the logic.

From an approved cell to a production order

Size the order from your production, not from the pack list: wicks = candles × wicks per candle × (1 + allowance), with 5% as our worked-example allowance for rejects and retests. Take 160 candles a month with C2 at 9% as the worked example.

Reorder point = daily use × your lead time + safety stock. At 168 wicks a month you use about 5.6 a day; with a 12-day order-to-bench time (your figure; CSI dispatches from Pune and publishes no lead times or stock depth) and a week of cover, reorder at about 106 wicks. Per-SKU reorder points across several wicks are the reorder-point post.

What to buy, and what to leave on the shelf

Five items, chosen on page statements, price and stock on 1 October 2026. CSI publishes no wick-performance data, so nothing here is ranked on burn results. Oil candle counts use 13.5 g per 150 g candle at 9% and 12 g at 8%.

Leave on the shelf for this problem: Performance Wooden Wicks (₹590.00 for 10): one listing with no widths, so it cannot step up a rung; wood against cotton is a separate test. Eco Braided Wicks (₹472.00 for 5 m): unsized, cut to length; only worth a look if C1 and C2 bracket the answer between them.

Frequently asked questions