My Replacement Fragrance Performs Well at 6% While the Original Required 9% — Should I Redesign the Formula Around the Lower Effective Load?

Why a lower load is a redesign, not a discount

CSI calculates fragrance load as a share of the finished candle: oil = W × L, and the wax is everything else. So when a 200 g candle moves from 9% to 6%, the oil falls from 18 g to 12 g and the wax rises from 182 g to 188 g. Nothing else in the jar changes, but the melt pool is now a slightly different liquid — more wax, less oil — and the wick is feeding on that different liquid for the whole life of the candle. In general candle-making practice, fragrance load is one of the things that can influence how a wick burns, how the melt pool behaves and how the surface looks after storage. That is why the lower load is a formula change, even though it looks like a saving on the purchase order.

The reader's figures in the title — 9% for the original and 6% for the replacement — are used exactly as given. We do not claim that any CSI oil performs at either load; CSI publishes no throw data for any oil. To make the costing concrete, this page uses an illustrative pairing: Neroli Fragrance Oil as the original and Neroli Signature Grade Fragrance Oil as the replacement. They are two different oils with two different prices; whether the second one would work at 6% in your candle is exactly what your test has to find out.

There are really three questions hiding in the title. First, is 6% genuinely good, or only good on the first burn, in the first week? Second, is 6% the lowest load that passes, or would 5% do? Third, does anything in the rest of the candle — wick, jar heat, storage — behave differently at the new load? A redesign answers all three. A discount answers none of them.

What the lower load is worth, per candle and per batch

Cost per candle at a given load is simply price per gram × W × L for the oil, plus the wax that fills the rest. With the 1 kg listed prices, Neroli Fragrance Oil is ₹6.66 a gram and Neroli Signature Grade is ₹7.60 a gram — the replacement is the dearer oil per gram. It still wins at 6%, because it uses a third less oil. Luxury Soy Wax CSI 464 is ₹507.40 a kilo at 1, 5 and 10 kg, so every gram of oil you remove adds back ₹0.51 of wax.

Two things stand out. Keeping the old 9% with the new oil would have made the candle dearer by ₹16.99 — so "just swap it in at the same load" is not only untested, it is the most expensive option. And the saving from the redesign is ₹25.56 per candle after the extra wax is paid for. The extra wax is small (₹3.04), but leave it out and you overstate the saving and, worse, under-order wax. The general case — two oils, two prices, two loads — is worked in the ₹5,000 vs ₹3,500 per kg page.

The break-even point is worth knowing. With these prices, the replacement costs the same as the original at about 7.80%. Below that, it saves; above it, it costs more. So if your bracket test ends up choosing 7% rather than 6%, the redesign still pays — but the margin halves. If it drifted back towards 9%, it would stop paying at all. That is the commercial reason to test the lower rungs properly rather than settling for the first load that seemed fine.

The retest matrix for a lower-load redesign

The verdicts below assume the wax, wick, jar and dye stay exactly as they were and only the oil and its load change. KEEP means still valid, CONFIRM a short side-by-side check, REPEAT a full retest, NEW a test your old formula never needed.

Notice what is not on the KEEP list: the wick. The instinct is that a lighter load is gentler, so the old wick must be fine. It may well be — but it is a different melt pool, and the only evidence that counts is a burn against the control. The halfway-down page shows why the late burns are where differences tend to surface, and the oud page covers the opposite situation, where a new oil at the same load changes the melt pool.

The bracket test: proving 6% is the right number

The design below is a working practice; change it to fit your range. It brackets the proposed 6% with 5% and 7%, three candles each, and adds a single candle of the replacement at the old 9% — not because you intend to use it, but because it tells you whether extra oil buys anything at all. If the 9% candle is no better than the 6% ones, you have strong evidence the lower load is not leaving throw on the table. Three control candles of the original at 9% are poured on the same day from your remaining stock.

The scoring sheet

Every line is recorded against the same-day control. Throw is scored 1–5 relative to the control; everything that can be measured is measured. Consumption in grams per hour is weight lost divided by hours burned — useful here because a leaner oil load can change how fast the candle is consumed, and that changes the burn time you can honestly claim.

CSI publishes no numeric flame or jar-temperature limit and this sheet does not invent one. Safety lines pass when they are no worse than the approved control on every burn and stay within your jar supplier's guidance and the fire-safety standard you work to (ASTM F2417 is a commonly cited one). A cheaper recipe that runs hotter is not cheaper; it is a different risk.

The priced bracket kit

The new oil for the bracket is 126 g — three rungs of three candles plus the single 9% check — and with a 5% allowance for weighing and pour loss (a working assumption) that is 132.3 g. The wax covers every candle including the controls, with the same 5%. The controls' oil comes from your remaining original stock and is not priced here.

At a saving of ₹25.56 per candle, the kit pays for itself in 152.5 candles — one-off revalidation cost divided by the saving per candle. Your own labour, electricity and the time the candles spend curing are real costs too; add them as your own example figures (for instance, a day of your time at whatever you value it) and the payback stretches, but on any production range it is still measured in a few hundred candles, not thousands. At an example volume of 300 candles a month (replace with yours), the redesign is worth ₹7,668.36 a month before labour.

The decision rule and the release record

Write this rule before the bracket is poured: adopt the lowest rung whose cold throw, early hot throw and late hot throw all reach the control, with every safety line no worse than the control through a full-life burn, and no sweating or colour problem in storage. If 5% passes, 5% is the redesign — not 6%. If only 7% passes, redesign at 7% and recalculate the saving. If nothing below 9% passes, the "6% performs well" observation was a first-week impression, and the oil stays at the old load or is rejected on cost.

Once a rung is adopted, write it into your formula file as a revision of the SKU: the replacement oil, its order date and bottle size, the new load, the unchanged wax, wick, jar and dye, the tests run and the date. Keep a sealed 15–20 g retained sample of the replacement oil and one unburned candle from the first production batch. CSI does not publish lot numbers or change notices, so your own sample is the reference if a later bottle ever seems different — the reformulation page explains how to use it. And request the replacement's MSDS and IFRA documents on WhatsApp for your file.

Where the neighbouring pages take over

This page is about a replacement that needs less oil. If the replacement needs more, or simply behaves differently in wax, the lavender page sets up a full load ladder. If it is noticeably stronger and you are tempted to cut the load by feel, the sandalwood page is the one to read first. If a changed load has started to sweat, the sweating page takes the lower-load route before any wax change. For the pure economics of two oils at two prices, the ₹5,000 vs ₹3,500 per kg page works the break-even load in general. And if you are doing this work in advance, before a stock-out forces it, the backup-supplier page shows how to pre-qualify replacements on your own timetable.

Frequently asked questions