The Wick: The Invisible Engine of a Candle
What burns in a candle is not the wick itself. The wick is simply a pump. Heat melts the wax, and the molten wax (fuel) travels up through the wick's fibres by capillary action, vaporises at the base of the flame, and that is where it actually burns. The wick's job, in a single sentence, is to deliver exactly as much fuel as the flame consumes — no less, no more.
When that balance is lost, failure occurs in one of two directions. Too little fuel and the flame shrinks, heat cannot spread across the surface, and tunnelling begins. Too much fuel and the flame grows, carbon cannot combust fully, and you see soot. The majority of complaints candle makers receive stem not from the fragrance but from this single balance going wrong.
Then there are the loads to consider. The wick must carry not just pure wax but also the fragrance oil, dye and any additives you have incorporated. As the fragrance oil load increases, the melt pool thins or thickens, its viscosity changes, and the chemistry of the mixture the flame consumes shifts accordingly. Wick selection is therefore not a standalone decision; it is the combined output of the wax type + vessel diameter + fragrance load + dye quartet.
The fundamental rule: the wick is chosen for the complete formula. If you have changed the fragrance, increased the rate from 8 % to 10 %, or deepened the dye colour, the wick in your hand is no longer "validated". Burn testing starts from scratch.
We said earlier that in perfumery, when the top notes fade, the next layer takes the stage; candles follow a similar handover. The scent you perceive from an unlit candle (cold throw) is the work of the packaging; the scent that fills a room when the candle is burning (hot throw) is directly the work of the melt pool the wick creates. No pool — no scent.
Wick Selection by Wax Type, Vessel Diameter and Fragrance Load
The numbers in wick catalogues are not a universal standard. Every manufacturer's braid, cotton-to-linen ratio and optional inner core differs. Read the ranges below as a starting point, not as a formula.
First, establish your wax type. Paraffin is a low-viscosity, easily vaporised fuel; for the same diameter it generally requires a smaller wick. Soy and other plant-based waxes have a higher melt viscosity, which makes capillary draw harder; they typically need a wick one or two sizes larger. Coconut–soy blends are soft and open a melt pool quickly but can cause adhesion issues at the vessel wall. Beeswax is the most demanding: its high melting point and dense structure call for the thickest wicks.
| Wax type | Typical fragrance load range | Wick tendency | Key consideration |
|---|---|---|---|
| Paraffin (container candle) | 6–10 % | Finer / smaller size | Higher soot tendency; wick length critical |
| Soy (container candle) | 6–10 % | 1–2 sizes larger | Wall adhesion, tunnelling risk, matte surface |
| Coconut + soy | 8–12 % | Medium–large | Soft structure; pool overflow at high loads |
| Beeswax | 3–6 % | Thickest / usually square braid | Natural scent can mask fragrance oil |
| Wax melt (wickless) | 8–12 % | — | No wick; all performance depends on load and wax base |
Ratios are given by weight; filling is done by volume. Always weigh fragrance oil in grams as a percentage of total wax weight. The specific gravity of fragrance oils varies widely (citrus-forward compositions are around 0.84, while heavy resin/synthetic-heavy ones can exceed 1.10). "10 ml of fragrance oil" could be 8.4 g or 11 g — at a 10 % load that difference directly affects wick behaviour. Always use a scale.
Diameter is the second determining factor. The vessel diameter that can be reliably managed with a single wick is generally up to around 7–8 cm. Above that, a single wick either cannot open the pool (tunnelling) or must be oversized to do so and produces soot. For wide vessels, the correct solution is not one large wick but multiple wicks.
| Vessel inner diameter | Number of wicks | Placement | Notes |
|---|---|---|---|
| 5–6 cm | 1 | Dead centre | Oversized wick in a small diameter = glass overheating |
| 6.5–8 cm | 1 | Dead centre | Most common vessel; most test variations occur here |
| 8–10 cm | 2 | Symmetrical about centre, ~1.5–2 cm from each side | Each wick selected smaller than a single-wick equivalent |
| 10–13 cm | 2–3 | Triangle / opposing | Watch for flames feeding each other |
On wick families: flat cotton-braid wicks (CD, ECO and similar types) are common in plant-based waxes; they curl slightly and partially self-trim. Square braid types are preferred for beeswax and dense fuels. Cored wicks stand upright and do not topple in a large pool. Wooden wicks produce a wide, flat flame with a pleasing crackling sound but have a narrow thickness/cross-section tolerance — choose incorrectly and they extinguish easily.
Increasing the fragrance load does not automatically mean "go up a wick size". Some fragrance oils thin the melt and accelerate capillary draw — in that case a larger wick will produce soot directly. Others weigh down the pool and starve the wick. The direction depends on the fragrance's chemistry; the only reliable answer is a burn test.
Tunnelling, Soot and Drowning: Assign the Fault to the Right Cause
A maker's first instinct is usually to blame the fragrance oil. Yet for all three classic failures described below, the primary cause is the wick; the fragrance is only a secondary variable.
Tunnelling
The candle bores downward from the centre, leaving a thick wall of unmelted wax around the edge. There is a single cause: the melt pool is not reaching the vessel wall. Wax develops a "memory ring"; wherever the pool reached on the first burn, it tends to stop on subsequent burns. If the wick is too small or the user cut the first burn short, tunnelling is inevitable.
Work through solutions in order: first increase the wick size incrementally, then question the vessel diameter, and only lastly the wax base. In soy candles the tunnelling risk is noticeably higher than in paraffin.
Soot and carbon mushrooming
If a carbon ball is building up on the wick tip like a mushroom cap, smoke is visible, and the glass is blackening, the flame is drawing more fuel than it can combust. First check wick length: a wick not trimmed to approximately 5–6 mm before each burn will produce soot even at the correct size. If the length is correct, the wick is oversized. A high fragrance load and dark dye amplify this tendency, but the fragrance alone is rarely the sole cause.
Drowning (self-extinguishing)
The flame sinks into the pool, diminishes and goes out. The wick has become submerged under more melt than it can handle. Causes: wick too small, pool too deep (depth exceeding 10 mm is a warning sign), fragrance load beyond the wick's capacity, or insufficient oxygen because the vessel opening is too narrow.
| Symptom | Primary cause | First action | Relation to fragrance |
|---|---|---|---|
| Wax wall at edge / tunnel | Wick too small or first burn too short | Go up one wick size; instruct the customer on the first-burn rule | Indirect: heavy load makes draw harder |
| Black smoke, glass darkening | Wick too large or too long | Trim to 5–6 mm; if problem persists, go down a size | Indirect: high load + dark dye increase tendency |
| Carbon ball on wick tip | Excessive fuel flow | Switch wick type (self-trimming braid) | Some heavy molecules contribute |
| Flame extinguishing | Pool too deep / wick capacity insufficient | Go up a size; keep pool depth below 10 mm | Direct: excessive load drowns the wick |
| Oil pooling / sweating on surface | Fragrance load exceeds wax base's binding capacity | Reduce load or change wax base | Direct: this one genuinely is fragrance-caused |
| Flame too tall, flickering | Wick too large or air draught | Go down a size; shield test area from draughts | Unrelated |
| Weak scent, clean burn | Load too low or composition not suited to hot throw | Increase load incrementally; evaluate the fragrance | Direct |
The last row matters. If the burn is flawless, the pool is full and there is no soot but the scent is not filling the room, stop adjusting the wick — the issue is fragrance load and the composition's volatility profile. A composition dominated by heavy resins/musks and not designed for candles may deliver a weak hot throw even from a perfect melt pool.
In candle making, the flash point (flash point) of the fragrance cannot be overlooked. If you add the fragrance oil when the wax is excessively hot, you will lose some of the volatile top notes. General practice: add the fragrance at the blending temperature recommended by the wax supplier and stir homogeneously for at least 2 minutes. Good ventilation during melting, no open flames, and wearing gloves and eye protection are standard requirements.
Burn-Test Protocol: Replace Guesswork with Data
Wick selection is found through experimentation, not catalogues. But experimentation must itself be set up with discipline. The protocol below is the shortest path to comparing three or four candidate wicks for a single candle formula.
- Establish the single-variable rule
Same wax, same fragrance oil, same rate, same vessel, same pour temperature. The only thing that changes is the wick size. If you change two things at once, you will never know which one caused the result. This is the candle-making equivalent of a base compatibility test mindset.
- Define your candidate range
Start with the size one below the supplier's recommendation, the recommended size itself, and one above. Pour at least two candles per candidate — one for burn testing and one for comparison and scent evaluation.
- Cure the candles
After pouring, store the candles covered, at room temperature, away from light. For paraffin, 24–48 hours is usually sufficient; for soy and other plant-based waxes, a curing period of 1–2 weeks makes a noticeable difference to scent development. Tests carried out on uncured candles are misleading.
- Run burn sessions systematically
Trim the wick to 5–6 mm before every session. Burn in a draught-free room at around 20–22 °C on a flat surface. Session duration: approximately 1 hour per 2.5 cm of vessel diameter; generally 3–4 hours. Do not exceed four hours.
- Record five data points every session
(1) Time to achieve a full melt pool, (2) pool depth (mm), (3) flame height, (4) presence of carbon mushrooming / soot, (5) outer glass surface temperature. A test conducted without a notebook is an impression, not a test.
- Continue until the candle is finished
Most faults appear not during the first session but after the candle is half consumed. As the flame descends deeper into the vessel, oxygen decreases, the glass heats up and drowning begins. Carry the test through to the last 1–2 cm.
- Check safety thresholds
If the outer glass surface becomes too hot to touch, the wick is too large — the candle cannot be sold to a consumer. If the flame exceeds 3–4 cm, flickers or splits, the same decision applies. If there is any risk of the glass cracking while the candle continues to burn near the base, a warning label indicating the remaining safe burn height is mandatory.
- Validate the winner
Pour a new batch with your chosen wick and repeat the test. A wick that works in a single candle does not always give the same result across a production batch or with a different fragrance oil.
Suggested record format: keep a table with one row per session — date, wick code, fragrance and rate, session number, pool diameter/depth, flame notes, soot notes. When a customer complaint arrives six months later, you will have records rather than guesses. These records also protect you on the product notification and liability side.
Measure scent performance separately from the burn test. For cold throw, leave the candle in a closed room for 20–30 minutes and evaluate it after removing the lid. For hot throw, burn the candle in a room of around 10–15 m² for 2 hours, leave the room, and return 10 minutes later — an evaluation made by a nose already acclimatised to the scent is worthless.
Frequently Asked Questions
The most common questions from the workshop, with concise, actionable answers. The rest is down to you.
How long should a candle wick be left?
Can tunnelling be fixed after it has started?
My candle is producing soot — is the fragrance to blame?
Which is better for a soy candle — a cotton braid wick or a wooden wick?
If I increase my fragrance load from 8 % to 10 %, should I go up a wick size?
When is a double wick necessary?
My candle keeps going out while burning — why?
The burn looks perfect but the scent is not filling the room — is the wick too small?
What happens if the wick is not centred?
How long does burn testing take?
Do dyes and micas affect the wick?
I make wax melts — do I need to worry about wick selection?
Related Articles
How to Perform a Fragrance–Base Compatibility Test: Shampoo, Soap, Cream, Candle and Room Fragrance
Testing fragrance oil in the target base at the real dose and conditions: the challenging characteristics of five bases, typical problems and control samples.
Read →Candle & Wax Melt: Fragrance Load, Hot/Cold Throw, Paraffin vs Soy
Load calculation and material selection for candle scenting.
Read →Fragrance Oil in Soap: Cold/Hot Process and Usage Rate
Scent stability in soap, skin deposit and usage rates.
Read →