If One Fragrance Consistently Creates Poor Adhesion, Should I Change Its Load Rather than Changing Wax Across My Entire Collection?

Change the oil's load or the collection's wax? The answer

When one fragrance consistently pulls away from the glass and the rest of the collection does not, the evidence is pointing at a single SKU. Fragrance oil is commonly described as changing how soy wax crystallises and contracts, and each oil is a different mixture, so it is entirely plausible that one oil sits closer to the adhesion edge in your wax and jar than the others. The proportionate response is an SKU-level fix: find the highest load of that oil that stays on the glass and still throws, and leave every other SKU — with its validated wick, load, cure and throw — exactly as it is.

Changing wax across the entire collection is the opposite: it treats a one-SKU symptom with a collection-wide cure. Every SKU in the new wax becomes a new candle, so wick, fragrance load, cure and hot throw all have to be re-validated, and adhesion has to be re-screened for every oil, because an oil that is fine today in your current wax has not been proved fine in the next one. All three CSI container waxes make the same adhesion claim on their product pages, so nothing printed on a page tells you the new wax will fix the difficult oil either. The wax switch might be right one day; it is not the first move.

This page works the decision through an example: a 10-SKU collection poured in CSI's Borosilicate Glass Jar 600 gram at an illustrative 550 g fill (the page says "ideal for 500–600 g wax fill"; weigh your own), in Luxury Soy Wax Chunks at 8%, where the difficult oil is Roasted Coffee. Path A is a twelve-jar load ladder for coffee alone. Path B moves all ten SKUs to Premium Coconut Soy Wax CSI 468. The numbers below are arithmetic on CSI's listed prices; the collection size, monthly volumes and reject rate are labelled examples you replace.

First, is it really that one fragrance?

"Consistently" is the key word in your question, so check that the consistency is real. Has the difficult oil spotted across several batches, on different days, in jars from different deliveries, beside other SKUs that stayed clean in the same session? If it is always poured last, or always in the biggest batch, or always in the darker colour, those habits travel with it and may be the real cause. The two-day replicated pour is the clean way to prove an oil difference, and the section hub's split pour separates the oil from the scented routine.

Also confirm the load really is what the recipe says. In a 550 g candle at 8%, CSI's convention (oil = W × L) puts 44 g of oil in each jar and 506 g of wax. A recipe written as "8% of the wax" puts in 40.5 g instead — close, but a recipe rounded up "because coffee needs more" can be carrying a good deal more than its siblings. And check the defect is wet spots, not sweating on the top surface or a colour change seen through the glass; the dark-patches post helps you tell them apart.

Path A: the load ladder for one oil

The ladder runs the difficult oil at its current 8% and at two lower loads, beside an unscented control, all from one melt of your current wax into one delivery of production jars. In Luxury Soy Wax Chunks, CSI's product page states a fragrance load of up to 13%, fragrance addition at 80–90°C, pour at 75–80°C in its spec table (the step list also shows 70–80°C), a two-minute stir and a cure of at least 48 hours, ideally one to two weeks. The Borosilicate jar's page publishes no hot-pour limit; it does claim a "smooth inner surface for … clean adhesion" — a claim to test, not evidence — so check your supplier's guidance before pouring at the top of a wax's range.

Because a 550 g candle is a large mass of wax, it cools slowly and its glass stays warm for longer than a small jar's, which is one reason results do not transfer between jar sizes. Run the ladder in the jar you sell. If your production run uses a different jar, container shape inside wax qualification explains why the ladder must move with it. Optimising load before replacing wax covers the ladder method in more depth for a smaller jar.

Path B: what a collection-wide wax switch really involves

It helps to see Path B written out with the change-control vocabulary this series uses — KEEP (result still valid), CONFIRM (short side-by-side check), REPEAT (full retest), NEW (a test never needed before). A wax change is the biggest change a container candle can have short of a new jar: it re-opens almost everything.

Translated into jars, Path B for the example collection is 10 SKUs × (3 adhesion jars + 3 wick and throw jars) + 12 unscented controls over 4 pour days = 72 jars. In this example's Borosilicate jar at ₹200.00 each, the jars alone are ₹14,400.00 — and CSI's listing shows the Borosilicate 600 g jar in limited stock, so confirm on WhatsApp before planning anything beyond a test quantity. Add 3 × 10 kg + 1 × 5 kg + 4 × 1 kg of CSI 468 (₹24,895.00), wicks (₹531.00) and stickers (₹200.00): ₹40,026.00 in CSI materials. Your own oils add another 2,772 g — valued, purely as an example, at Roasted Coffee's 1 kg rate, about ₹10,361.74.

The money: development cost, then cost per candle

The two paths differ twice: once in development, once on every candle afterwards. At production rates — Chunks and CSI 468 at their 5 kg rates, Roasted Coffee at its 1 kg rate, the Borosilicate jar at ₹200.00 and an Eco Thin C1 at ₹5.90 as the example wick line — the example candles cost:

Path B's wax line rises by ₹19.23 a candle because CSI 468 is ₹637.00 a kilo at the 5 kg rate against ₹599.00 for Chunks. On an example output of 300 candles a month across the collection, that is ₹5,768.40 a month — ₹69,220.80 a year — paid on the nine SKUs that never had a problem. Path A moves the other way: if the coffee SKU (30 candles a month in this example) passes at 6%, it saves ₹34.53 a candle, ₹12,430.44 a year. And if nothing is done, an example 25% of coffee jars spotted costs ₹5,050.99 a month in materials sitting in jars you must hide, discount or re-pour. Every figure in that paragraph except CSI's prices is an example: use your own volumes and counts. Whether a dearer wax pays for itself and including development cost in a switching decision take this further.

The decision rule

Read the ladder at 7 days and after the temperature cycle, then run the throw test on the clean arms. Stop at the first line that is true.

Two waxes in one workshop is a real cost — two bags to stock, two pour routines, two sets of labels on the melter — and some makers prefer to replace a difficult oil rather than run an exception. That is a business choice, not a technical one; the ladder and the single-SKU wax arm give you the evidence to make it. Fragrance-specific screening makes sure the next difficult oil is caught before its first batch.

The CSI products in this example

Where to go next

This post owns the scale of the fix: one SKU or the whole collection. The load post owns the ladder method; the replicated pour owns proving an oil difference; the screening post owns catching the next difficult oil early; the dye post covers colour as the other ingredient that travels with a SKU; which tests repeat after a blend or wax change lists the work a Path B really triggers.

Frequently asked questions