How to Test Fragrance Oil for Discoloration Before Bulk Production

What a colour failure costs at bulk scale

A slow colour change does not show up on pour day. It often shows up weeks later, on your shelf or in a customer's photo, by which time the whole batch is poured, labelled and possibly sold. At that point there is no fix: you can re-describe the candles, move them to a different line, or discount them. The test exists to move that discovery from week six of production to week four of testing, while the only thing at stake is a set of test jars.

The protocol's materials are 13.1% of the Palo Santo batch, and they test two oils at once. The figures exclude labour, freight, packaging and GST; add those and a failed batch costs more still. A batch that fails on colour also costs something that does not appear in any column: the trust of customers who received a white candle that turned beige. If your plan is 5 kg rather than 1 kg, the argument is five times stronger. The wider checklist before a 1 kg order, covering throw, wick, sweating and stock depth as well as colour, is in What Should I Test Before Buying 1 Kg of a New Candle Fragrance Oil?.

Is it the oil? Design the test so it can answer

Colour in a white candle has several possible causes: the oil's own colour, vanillin or similar components, a trace of dye, heat, the wax's natural shade and light. A pre-bulk test is only useful if it isolates the oil from all the others, and the design below does that with four elements, each of which answers one "is it actually the oil?" question.

Everything else is held constant, and that matters as much as the controls: one wax lot and one melt for all twelve jars, one jar type, one wick type, loads weighed on a scale, add and pour temperatures read with a thermometer and written down. If your production candles are dyed, dye the test candles identically, and add a dyed unscented control, because dye is then part of the system being tested. If you have not yet ruled out day-0 causes such as the oil's own colour, a four-jar check in My White Candle Turns Beige After Adding Fragrance Oil — Why? is quicker and should come first.

The 30-day protocol, step by step

This is deliberately more formal than a curing log. If you only want to watch how one existing batch changes while it cures, the lighter method with a half-foiled jar is in Can Fragrance Oil Change Candle Colour During Curing?; if you are still choosing between fragrance families, the five-jar sampler in Best Fragrance Types for Minimal Discoloration in Candles comes first. This protocol is the last gate before money goes into bulk.

The photo log: making day 30 comparable with day 1

A photo log is only evidence if every photo is taken the same way. A common way for a test to go wrong is not the oil but the pictures: morning daylight on day 1, a warm evening bulb on day 30, and a "yellowing" that is really a change of lamp.

The five-step scale is simple on purpose. 0: indistinguishable from the unscented control in the same lane. 1: just visible when placed beside the control. 2: clearly warmer beside the control, visible at arm's length. 3: obviously cream, beige or straw without a comparison. 4: yellow, honey or brown. Rate the jars in the room, side by side, as well as from the photos; the photo is the record, your eyes are the judge.

Pass, conditional pass, fail: rules before day 0

These are the rules we would use for a white or undyed candle in clear glass. They are editorial judgement, not an industry standard; tighten or loosen them for your range, but do it before day 0 and write them on the log sheet.

Two notes on reading the table. First, the dark lane is the one that decides whether an oil can be used in white at all, because that is the oil changing on its own; the light lane decides how you must store and display it. Second, a fail is not a verdict on the oil's quality or throw. An oil that fails for white may be a perfectly good oil for an amber-jar line; the policy for that is in Should I Avoid Vanilla Fragrances When Making White Candles?, and the costs of presenting a darker candle well are in My Vanilla Candle Has Excellent Hot Throw but Turns Brown — How Do I Keep a Premium Appearance Without Losing the Fragrance?.

What the protocol costs

Each candidate uses 48 g, which leaves 2 g in a 50 g bottle: not enough for a rerun. If you expect to repeat, buy 100 g of a candidate you are serious about. Oil cash alone is ₹742.30. The wax and the eight spare jars are not wasted; the scented test jars are usable candles once the verdict is in, except that you should not sell a jar from the light lane as new stock. All figures exclude labour, freight, packaging and GST.

What to buy for the test

The vessels first, then two worked-example candidates that show two situations you will often meet (a page that states 0% vanillin, and a page that says nothing), then the positive control. Swap the candidates for your own; the design does not change. Oils are chosen on page vanillin wording, stated notes, a stated candle load inside CSI 464's 10% ceiling, price per gram and stock at every size on 25 September 2026. CSI has no colour measurements for any of them. Candle counts assume 150 g at 8%.

After the verdict: the bulk decision and the calendar

A colour pass is one gate, not the whole decision. Before 1 kg, the oil also needs a load that throws in your jar, a wick that gives a full melt pool, no sweating after two weeks, and a cost per candle your price can carry; a full qualification checklist for a new oil in a candle business is in How to Qualify a New Fragrance Oil Before Adding It to Your Candle Business. When everything passes, order the bulk size with the batch details written down, because a re-order months later should be tested against the same controls if the colour matters to you.

Now the calendar, honestly. Diwali falls on 8 November 2026, 44 days after these prices were checked. A 30-day test poured by 1 October gives a verdict on 31 October, which leaves no room for a cured production batch before Diwali. For a Diwali white range, use an oil you have already logged, or accept a day-14 interim read with a stricter rule (dark lane at 0, light lane at 0 or 1) and box every candle. Christmas, on 25 December 2026, is 91 days away and fits the full protocol with time to spare.

Frequently asked questions