My Candle Smells Strong During the First Few Burns but Becomes Weaker as It Gets Halfway Down the Jar — Why Would Fragrance Performance Change Through the Candle?

What changes as the wax goes down

A container candle is mixed once and poured once, so, barring a problem in the pot, the wax at the halfway mark holds the same oil at the same load as the wax at the top. What changes as you burn down is everything around that oil. Here are the changes, each described as a tendency, not a certainty.

The pool's size, depth and position

Hot throw comes from the melt pool, and the pool is shaped by the jar. The 150 g jar used in the worked example has a "curved silhouette", an outer diameter of 7 cm and a base diameter of about 6.5 cm on its page. If the inside narrows by about half a centimetre over the burn, a full pool loses about 13.8% of its area: π × 3.5² is 38.48 cm², π × 3.25² is 33.18 cm² (worked example; no CSI listing states an internal diameter, so measure yours). A wick that leant as the wax set can give a pool that is round at the top and lopsided lower down, with a hang-up wall on one side. And a pool halfway down sits several centimetres below the rim; the glass above it becomes a chimney the vapour has to rise through, a geometry question that the vessel-swap test owns.

The heat, and the flame

As the level falls, the glass above the pool warms and shelters the flame, the pool tends to deepen and the flame can grow; in other candles a wick at the low edge struggles as the pool deepens around it. Either can change throw. A candle that gets stronger near the bottom is a different warning and belongs to the too-strong-at-the-bottom post; one whose wick drowns in the last third is late wick failure. This page is about the gentle, mid-jar fade.

The wax's burn history

Every burn melts a pool, and when you blow it out that pool sets again with less of its lightest material, because top notes tend to leave a hot pool first. The next burn melts that wax again, plus a new layer beneath it. Long sessions keep more wax hot for longer, so by halfway down more of the candle has already been through the heat. That is why this page compares a capped-session candle with a four-hour one: if session length drives your fade, the two will part company.

The person and the room

By the middle of a jar you have smelled the same candle in the same room on many evenings. Olfactory adaptation is well established within one burn; over weeks, familiarity with one scent in one room tends to make it register less, which is why the log uses a fresh nose who does not live with it. The room changes too: monsoon humidity, the AC season, a fan switched on for summer. Record the conditions at every mark.

Depth patterns and the cause behind each

The table maps the shapes a quarter-mark log can show to the likely cause. The order is editorial judgement of how often we would expect each; CSI publishes no failure statistics and has no measured throw data. The right-hand columns are what to look for in the log to confirm or dismiss each one.

Patterns overlap. A maker who burns four-hour evenings in the same room can show pattern 1 and pattern 3 at once, which is why the log scores both your nose and a fresh one, and burns both a capped and a four-hour candle. Rows 1 to 3 are the heart of this page; rows 4 and 5 hand you to the post that owns the event; row 6 is the happiest result and is not rare.

The quarter-mark burn log

The log uses two candles from the same pour: same wax, oil, load, wick and jar, cured the same number of days. Candle A is burned in capped sessions: stop soon after the pool reaches the jar edge, and never beyond about four hours. Candle B is burned in four-hour sessions, the common-practice cap and the longest we would suggest; if your pool already takes close to four hours to reach the edge, the two schedules barely differ and you should read the too-little-heat post first. Burn them on alternate days in the same room, one at a time, within sight, away from draughts.

Bill of materials, if you need to pour the pair

Unburned twins from the batch that fades are best. If you have none, pour two from one pot in one session: at 8% of total weight each 150 g candle is 12 g of oil and 138 g of wax, so the pair takes 24 g and 276 g. With CSI 464 in the 150 g jar, pour at 80–85°C, and cure at least the 48–72 hours the Citrus Lemon page states for soy; many makers wait a week or two before the first scored burn.

Costs are worked examples and exclude labour, freight, packaging and GST. Starting from nothing, the cash outlay is ₹1,680.40, most of it jars and wax you keep; the 500 g bag of CSI 464 (₹260.00) also covers the pair. The real cost of this test is time: two 150 g candles burned down in sessions, on alternate days, is weeks of evenings. That is why it is worth doing once, properly, with a helper.

Two logs, read side by side

Here is how two worked-example logs might read. The numbers are illustrative, chosen to show the shapes; your log is the data.

Both candles fall together, the fresh nose and yours agree, and the pool column tells the story: the wick leant, the pool went lopsided and the flame shrank. That is pattern 2 with a cause attached. The fix is at the pour (hold the wick upright while the wax sets) and, if the pool also narrowed, a wick check through the lower half. The oil is not in question.

Here the pool is full and centred at every mark, but B, the four-hour candle, falls from the quarter mark while A barely moves on the fresh nose: pattern 3, top-note depletion from long sessions, with B's deeper pool as a supporting clue. Your own scores fall on both candles, so pattern 1 is running as well: you have got used to it. Two causes, two free fixes: cap the sessions, and judge with a fresh nose. Nothing in this log asks for a new oil, wax or wick.

What each pattern asks you to change

For pattern 2, the jar's shape is part of the story. The container series covers how a jar that is wider at the top changes burn testing (wider at the top than the bottom) and which width to choose a wick for (top, middle or bottom diameter). For heat late in the jar, see why the final third matters; the glass is never judged by touch.

Money you don't need to spend

We sell the oils, jars, wicks and wax on this page. For a candle that is strong at the top and quieter halfway down, here is what we would talk you out of.

To be fair to the bigger sizes: the 1 kg of Citrus Lemon costs ₹8.00 less than two 500 g packs, and at 12 g a candle it works out at ₹33.98 of oil against ₹40.80 at the 50 g price and ₹36.00 at the 100 g. That is a production decision for a scent that has already passed a full burn-down, not a fix for this one.

If the candle also tunnels, sweats or smokes, do not try to fix everything from one log; the cluster hub orders the checks for weak throw, and the which-change-first ladder ranks the levers by cost once the evidence points at one.

Products for the depth log

Five items: the jar the log is built around, the worked-example oil, the wick step if the pool narrows, the centring tool if the wick leans, and the wax for a fresh pair. Chosen on what each page states, 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 12 g and log pairs at 24 g.

Frequently asked questions