My Candle Fragrance Performance Changes from Batch to Batch Even Though I Follow the Same Formula — What Raw-Material Variables Should I Investigate?

Why the swing usually starts at home

A formula is a set of numbers: 8% oil on total weight, CSI 464, a C1 wick, a 150 g jar. A batch is those numbers carried out with a particular bottle of oil at a particular point in its life, weighed on a scale in a particular state, poured on a particular day and judged at a particular age. Most of those change between batches without anyone deciding to change them. That is why the audit starts with your own stock and records: it is free, it is under your control, and in our judgement it is where most swings start.

It is also the honest place to start because of what nobody can see from outside. CSI does not publish lot or batch release data, composition or reformulation notices, and this page does not imply that any CSI batch has changed. If an order is the variable, the tools that can show it are your own: a sealed sample, a retained candle and a side-by-side test. When you need the lot number on an order, ask us on WhatsApp. For the wider question of which raw material to investigate first when a whole batch is weaker, see the same-formula, weaker-batch hub; for a new order that smells different in the bottle, the go/no-go gate is the new-order post. This page is the audit of what is already on your shelf.

The own-stock audit, ranked

The ranking is editorial judgement, ordered cheapest and quickest first, which is also roughly how often we would expect each to explain a swing; CSI publishes no failure statistics. You need your batch sheets, your bottles and your scale.

Cure age goes first because it costs nothing and is easy to get wrong: a batch judged at day 4 against one judged at day 15 is not a like-for-like comparison. Oil pages that state a cure say 48–72 hours for soy (25 of 102 in the text we captured; Dark Vanilla Oud's states none), and many makers judge at day 7 and 14. Compare batches only at matched ages, which is why the cure-standard post fixes a QC age for every batch. Process comes last here because it needs records you may not have kept; if you haven't, start today with the repeatable batch record.

The bottle map: which oil went into which batch

The diagram shows a worked example. A 500 g bottle of Dark Vanilla Oud is opened on day 0, and each batch of ten 150 g candles at 8% on total weight takes 120 g of oil (each candle 12 g of oil and 138 g of wax; 1,380 g of wax a batch). After batch 1 there are 380 g left and 24% of the bottle is air; after batch 2, 260 g and 48%; after batch 3, 140 g and 72%; after batch 4, on day 120, 20 g and 96%. Batch 5 takes those last 20 g plus 100 g from a new bottle: 17% of its oil came from a bottle open for five months, and the rest from one opened that day.

Why that matters, as tendencies rather than certainties: fragrance oils are blends of materials that evaporate at different rates, and each opening can let some of the lighter ones escape into the air space above the oil. A bottle that is mostly air, opened often, kept warm, over months, can drift in character or strength, and bright top notes are commonly the first to fade. We have no measurements of how much or how fast for any CSI oil, and will not invent a figure; the pair test below measures it for your bottle in your conditions. What the map gives you is a pattern to check: if your "weak" batches line up with the late life of a bottle, or with a batch that straddled two bottles, the bottle is a suspect before any supplier is.

Checking the scale with a known weight

A scale does not need to be badly wrong to cause a swing, and oil is the small number in the recipe, so it is where an error shows. Worked example: if your scale reads 5% high around 12 g, you weigh out about 11.4 g of oil when you mean 12 g; with the wax right, the real load is 7.65% instead of 8%. That is a modest difference, but batches weighed on different days, with a weakening battery or a scale moved to a new surface, can differ from each other in just that way.

The opened-versus-sealed bottle pair

If the audit points at the oil bottles, this is the controlled test. It compares oil that has been through your storage with oil that has not, and keeps everything else still. The cleanest control is a sealed bottle from the same order as your opened one (if you bought two 500 g, the second one). A bottle from a newer order also works, but then a difference could be your storage or a different lot, and this test cannot tell those apart; say so in your log, and use the new-order gate for the lot question.

Costs exclude labour, freight, packaging and GST; the oil from your own bottles is valued at the price you paid. Pour all four in one session at 80–85°C in the 150 g jar (CSI 464's page: 80–90°C; the jar's page: hot pour up to 85°C), judge them at day 7 and day 14, one candle per evening in the same closed room, alternating O and S, and score 0–3 on entering from outside. Before pouring, smell both bottles on a blotter strip, side by side and blind if someone can hand them to you: a bottle difference you can already smell is worth writing down.

Buying and storage habits that keep batches alike

If the pair says your storage is the variable, the fix is in how you buy and keep oil, not in a new supplier. These are judgement, not CSI data: buy in sizes you finish in weeks, not months for slow scents; keep bottles capped tight, cool and out of sun, away from windows, stoves and hot storerooms, which matters more in Indian summer (roughly April to June); decant a large bottle into smaller, full, tightly capped glass bottles when you open it, so most of the oil sits with little air above it; label every bottle with the date opened; and never let a batch straddle two bottles without noting it.

Buying size is where the honest arithmetic sits. For Dark Vanilla Oud, two 500 g bottles (₹5,690.00) cost ₹1.00 less than one 1 kg (₹5,691.00), and one stays sealed until you need it: the same money buys you a control for the pair test and less air in the bottle you are using. For a scent you pour slowly, the 100 g costs ₹0.21 more per gram than the 500 g, which is ₹2.52 a candle at 8%; in exchange, you never pour from a bottle that has been three-quarters air for months. Whether that is worth it depends on how fast you use the scent, and only your bottle map can say.

When your audit and pair point away from your own stock, and a new order seems to be the difference, message us with your order dates and ask for the lot numbers. We cannot give you composition or release data, because CSI does not publish them. If you are qualifying volumes, the repeat-order checks in the 5 kg gate are the right next step.

What the audit may need

Four items, most of them only if your own fail a check: a scale, the worked-example oil (as two 500 g rather than one 1 kg), a thermometer for the process line of the batch sheet, and wax for the pair. Chosen on page statements, price per unit and stock on 26 September 2026; CSI has no measured throw data, so nothing is ranked on throw. The oil table shows every size, 150 g candles at 8%, and how many ten-candle batches (120 g each) a bottle covers.

Frequently asked questions