My Candle Was Stored Near a Sunny Window and Now the Fragrance Seems Different — Could Heat and Light Both Be Contributing?

The verdict on the sun-side units

Start with dispatch, because that decision is the one with a deadline. The sun-side units are a stratified sample of their own: they had a different storage history from the units behind them, so inspecting two from the middle of the batch tells you nothing about them. Pull a sample specifically from the lit row, and inspect it against the five physical gates: hairline cracks in the glass, a wick that has drifted until it contacts the wall, a wick that will not sustain a flame, free oil pooled around the wick base, and a jar that becomes uncomfortably hot to hold during a test burn. Any of those and the unit does not ship, and you widen the sample from that row rather than shrugging.

A scent difference by itself does not clear that bar. Neither does a colour difference on one side, though a one-sided colour change is a strong hint about which exposure was doing the work — an unevenly darkened candle that is darker on the face that met the light points at exposure rather than at anything happening uniformly inside the wax.

And treat the rest of the batch as untested rather than as fine. If your storage has a lit row, the right correction is not "check the lit ones" once; it is to stop having a lit row, and to sample by position from then on.

Why a windowsill is two exposures

A sunny window is doing several things to a jar at once, and they are not the same kind of thing. The warmth is a bulk condition: it reaches every gram of wax in the jar, arrives and departs daily, and gives the wax repeated cycles rather than a single event. Soy is a soft wax that contracts as it cools, so each cycle adds to the candle's thermal history — that much is uncontroversial material behaviour and it is one of the reasons a container candle that has been through a hot season is not the same object it was on pour day, whatever it smells like.

Light is a surface-and-depth condition instead. It arrives only from one direction, it reaches only as far into the wax as the glass and the wax let it, and it stops entirely if the glass is opaque. That asymmetry is the single most useful diagnostic feature of a light contribution: it produces one-sided effects. Warmth does not care which way the jar is facing.

What must not be added to this picture is a rate. There is no published figure for how fast anything changes under either exposure — CSI's product pages state no shelf life, no storage temperature and no storage instruction at all, on any of the 107 fragrance-oil pages — and there is no honest arithmetic that converts a warm lit week into a notional number of ambient months. The pages do state things you can use: scent note structures, usage percentages, discolouration positions on 62 of them, and a measured flash point on 18. None of those is a stability rating. So the useful output is the comparison, and the comparison needs positions.

The differential: heat, light, both, neither

Four explanations fit the observation "it was near the window and now it smells different", and a fifth fits it too — that nothing about the candle changed and your reference point did. The table gives what each would look like alongside the scent change and the observation that separates it. Read the right-hand column as the work; nothing in the left-hand column can be confirmed by smelling one candle.

The fourth row is not filler. Perceived strength through glass moves with the temperature of the jar, with how long the lid has been closed, with what you smelled a minute earlier and with how familiar the scent has become. A candle brought in from a warm windowsill is warmer than one taken from a cupboard, and warmth alone changes what reaches your nose. Until both have sat in the same room for hours, any comparison between them is measuring the rooms.

The four-position test and the masked jar

This is a method rather than a standard, and the counts are choices: two units per position because it lets you read one at a mid-point and keep one for the end, and a hold matched to how long your own stock actually sits. Keep the structure — one pour, one jar, four positions, a control — and adjust the rest.

What the layout costs

Eight candles is a small number until you price it, so here it is line by line, with the rate used and the listed packs you would actually buy. Every figure is from a CSI product page in September 2026.

Two units of that eight are the control, so the "wasted" part of the exercise — stock you deliberately do not sell — is ₹422.78. That is the whole price of being able to attribute an observation. Against it, the thing you are protecting is a storage habit applied to every batch you pour: if your shelf has a lit end and you keep using it, every future batch inherits the same uncertainty, and every future complaint has the same two possible explanations.

What the glass can and cannot fix

If the layout points at light, the container becomes a real lever, and it is worth being precise about what it does. Opaque glass blocks daylight from reaching the wax. Tinted glass reduces it. Clear glass does not. That is a physical fact about the container, not a performance claim about any product, and it is the only fragrance-stability benefit a jar can honestly be credited with. None of these jars slows warmth down in any way you should plan around, and none of their pages claims to.

Read the capacity column carefully, because it is where jar arithmetic goes wrong. The clear jar states a fill in grams. The matte and frosted jars state 200 ml, which is a volume, and 200 ml of soy wax is not 200 g. If you move a 150 g recipe into a 200 ml jar you are changing the candle's size, its headspace and probably its wick — which is three variables, not one. Weigh one filled jar before you convert a recipe, and treat the swap as a new formulation to be tested rather than a substitution.

What to buy to run it

Four purchases cover the layout and the follow-up comparison. Each card lists every pack CSI publishes with the per-unit rate, so you can see where a pack step changes the per-jar or per-kilo price and where it does not.

What changes next time

Three changes follow from a windowsill result, in ascending order of cost.

Storage first, because it is free. Stop using the lit row for stock. If the only shelf you have is lit, box the stock — a closed carton converts a lit position into a dark one at the cost of cardboard, and it is the reason position B exists in the layout. Store lids on, and store the retained samples in the same conditions as the inventory so that the control remains meaningful.

Sampling second, because it is nearly free. Once you know that position produces different outcomes within one batch, a random sample is the wrong sample. Sample by position: lit row, dark row, top shelf, floor. Record the position on the sheet next to the unit number. This is the same discipline that makes warehouse sampling honest, and it costs one extra column.

Container third, because it costs money and changes the product. If the light contribution is real and you cannot control the display environment — a retail shelf, a hamper in a shop window, a display at a market — then tinted or opaque glass is a legitimate answer. Price it properly: the White Matte Glass Jar at ₹70.80 is cheaper per jar than the clear jar at ₹94.40, but it states its fill in millilitres and is a larger vessel, so the wax per candle goes up and the wick may need to change. Re-test the new combination as a new candle, not as the old candle in new glass.

Frequently asked questions