Why Is Fragrance Oil Leaking From My Soy Candle?

Leak or sweat? They look similar and they are not the same fault

Sweating is fragrance oil that has come out of the wax and is sitting on the candle — a film or a scatter of beads on the cured top surface, inside the container, where nothing but the candle gets wet. Leaking is oil that has left the candle altogether: a ring under the jar, a slick on the lid thread, a dark patch on the box liner, a stain creeping up the paper label, a wet line along the bottom seam of a tin. Same cause list, very different consequences, and the fixes diverge quickly.

This page is about the second one. Surface sweating has its own diagnosis, and CSI already has two long guides on it (eight causes of candle sweating and why your candle is sweating fragrance oil) — read those if the oil never gets past the wax. If oil is reaching the outside of the container, keep reading: the load, the bind and the route out all have to be checked, in that order.

The distinction matters commercially. A sweating jar can be wiped, cured longer and often sold. A leaking jar cannot: by the time a customer opens the carton, the oil has usually taken the label, the card sleeve and whatever the parcel was packed next to with it. On a 150 g candle in a Clear Glass Jar with Golden Lid, the materials alone are ₹213.23 — ₹76.11 of CSI 464, ₹36.82 of British Rose at 1 kg pricing, a ₹5.90 Eco wick and a ₹94.40 jar. Fifty of them is ₹10,661.30 before packaging, and a leak does not just cost the candle, it costs the second courier run.

The sixth row catches more makers than any other. A pale patch where the wax has lifted off the glass looks exactly like the start of a leak and is usually just air — the same contraction that causes wet spots and sinkholes. Air is cosmetic. It only becomes a leak route when there is unbound oil available to run into it, which is why the load audit comes before the cosmetic fix.

The three routes out: wall, base and rim

Route 1, the wall gap. Soy contracts as it sets. Where it contracts unevenly it can lift off the container wall, and any oil the wax is not holding will run down that channel to the base. Adhesion is the whole defence here, and it is one of the things CSI’s container grades are sold on: the CSI 464 page claims excellent glass adhesion, the Luxury Soy Wax Chunks page claims excellent adhesion to glass and containers, and the Premium Coconut Soy CSI 468 page claims the same plus an ultra-smooth finish. Those claims describe wax poured and cooled properly. Adhesion you break with a cold jar or a draught is adhesion you paid for and threw away.

Route 2, the base and the seam. A glass jar is one piece: there is no hole and no joint, so oil appearing under a glass candle has come down the outside of the glass from somewhere higher, or has been sitting in the carton from a neighbouring jar. A tin is different. It is a formed can with a joint where the wall meets the base, and that joint is a mechanical seal, not a weld. CSI’s Tin Containers are listed at ₹70.80 each in the 85 gram pack of 12 and ₹96.76 each in the 135 gram pack; the Golden Aluminium Tin works out at ₹30.00 each in its pack of 12 and its page describes a snug-fit lid. Tins are good containers. They are simply the wrong place to test a load you have not yet proved.

Route 3, the rim and the lid. Fill a jar to within a few millimetres of the rim and you have removed the candle’s expansion room. Wax expands as it warms; a jar that was flush at 28°C on your bench can be over the rim at the top of a stack in the sun. Once wax and oil touch the rim they follow the thread, wet the lid and run down the outside — which is why so many "base leaks" on glass are really rim leaks that arrived at the bottom. The Clear Glass Jar with Golden Lid page gives 150 g as its fill capacity and 8 cm as its height; that fill figure is the number to design to, not the number to beat.

Four causes, ranked by how cheaply you can rule them out

Cause 1 — the candle is carrying more oil than the wax holds

This is the first thing to check because it is free to check and it is the only cause that makes every other fix temporary. Each CSI soy grade publishes a maximum fragrance load as a percentage of total wax weight: CSI 464 up to 10%, Luxury Soy Wax Chunks up to 13% with 8% given as a starting point and 10% as the recommended figure, CSI 468 up to 13%. Eco Soy CSI 400 does not publish a fragrance ceiling — ask on WhatsApp before you push it, and treat a conservative 8% as the working assumption until you have an answer.

Then read the oil’s own page, because it can set a lower limit than the wax. Lavender Fragrance Oil states 6–10% for candles. Dark Vanilla states 6–10%. The Gardenia Tuberose page gives soy up to 10%. Where the two differ, the lower figure governs. A 150 g jar of CSI 464 at 12% is 18 g of oil against a stated ceiling of 15 g — two grams over on the wax and three over on most oil pages, in a jar that will spend a week in a warm warehouse.

Cause 2 — the oil never bound into the wax properly

Oil that goes in too cool, or is stirred for a few seconds instead of the stated time, is physically present at the right percentage and chemically not incorporated. The Luxury Soy Wax Chunks page is specific: add fragrance at 80–90°C and stir slowly for two full minutes. Eco Soy CSI 400 asks for 85°C, within 80–90°C, and a gentle two-minute stir. The CSI 464 page gives 80–90°C as the working range for melting and pouring. CSI 468 publishes a 50–53°C melting point and a 58–60°C pour temperature and no separate add temperature — so do not carry an 85°C soy habit across to it; ask on WhatsApp instead of guessing. More on the mechanism in the add-temperature guide.

The tell is that the fault follows the jar, not the batch: one or two jars from a melt leak and the rest do not, because the last pitcher went in cool or the stir was hurried on jar eight. A ₹354.00 Pen Thermometer removes the guesswork; a Mini Electric Wax Melter at ₹2,360.00 removes the temperature drop that small pitchers suffer in an air-conditioned room, which is exactly what one unverified reviewer above describes fixing.

Cause 3 — the cooling that opened the gap

Pour into a cold jar on a cold night, or let a fan or an open Mumbai window blow across the tray, and the wax nearest the glass sets before the middle. That is how you get the wall channel in route 1. Warm the jars, pour at the stated temperature, and leave the tray alone on a level surface until the tops are firm — no fridge, no fan, no moving them to another room to look at them. CSI’s own guide to smooth, crack-free candles and the note on how long soy should cool before pouring cover the same ground from the finish side.

Cause 4 — heat after the candle left your bench

A candle that was dry for three weeks on your shelf and wet on arrival did not change formula in the van. It got warm. Soy softens well below the temperature at which it melts, which is general candle-making behaviour rather than a CSI figure, and a softer wax body both grips the glass less and releases mobile oil more readily. CSI publishes a melting point for exactly one of these grades — CSI 468 at 50–53°C — and no heat-resistance or soft-point figure for any of them. That is worth saying plainly rather than dressing up: if you need a number for a shipping specification, ask for it on WhatsApp.

The Indian heat problem: your bench, the van, the customer’s shelf

Between April and June a Delhi, Ahmedabad or Nagpur afternoon runs past 40°C in the shade, and a closed vehicle, a metal shutter godown or a parcel waiting on a loading bay all sit above ambient. Against that, CSI 468’s published 50–53°C melting point is not a comfortable margin, and the soy grades generally begin to soften before they melt. You cannot control the van. You can control how much unbound oil is in the candle when it gets in the van, which is why the load audit is the first fix and the packing is the second.

Monsoon is a different question, and it is worth being accurate about it: humidity in Mumbai, Kochi or Kolkata does not push oil out of wax. It ruins the packaging that a leak lands on, softens cartons, lifts labels and makes a small leak look like a catastrophic one. If your complaints spike in July, check whether the candles are leaking more or whether the same small leak is now destroying a damp label. The two need different money spent on them.

There is a storage version of the same problem on your own premises. Finished stock stacked against a west-facing wall, or under a tin roof in a first-floor workroom, is being heat-cycled every day: warm by afternoon, cool by night, warm again. Each cycle is a chance for the wax to move against the glass. Keep cured stock low, shaded, upright and lidded, and keep the airtight clip-top containers (₹229.00, 550 g fill) or lidded jars for the stock you are holding longest.

Festive timing makes this worse before it makes it better. Diwali 2026 falls on 8 November 2026, so candles poured by about 15 October finish a 14–21 day cure between 29 October and 5 November — comfortably out of peak summer, but into the season when everything ships at once and parcels sit longer at hubs. Pour the festive range at a load you have already proved through a warm-leg test, not at the load you are hoping to launch. CSI’s note on planning fragrance inventory before the festive season covers the ordering side.

The load audit: what your candles are actually carrying

CSI product pages state fragrance load as a percentage of total wax weight: oil grams = wax grams × load. For a 150 g fill that is 9 g at 6%, 12 g at 8%, 15 g at 10%, 18 g at 12% and 19.5 g at 13%. Two habits push makers over the line without them noticing. The first is measuring oil by volume — a capful, a splash, a 20 ml beaker — when every published limit is a weight. The second is topping up a jar that came out short by adding wax only, which lowers nothing, or by adding oil only, which raises the load on the whole jar.

Weigh both. A Candle Making Weighing Scale is ₹354.00 and stock is low, so any 0.1 g kitchen scale will do the job; what matters is that oil is recorded in grams on a batch card with the wax weight next to it. If you cannot say what percentage a leaking jar was poured at, you cannot clear or convict the load, and you will end up changing the wax, the jar and the oil at the same time.

Look at what the over-pour is buying. Going from 15 g to 18 g of British Rose in a 150 g jar costs ₹9.20 more per candle and ₹460.20 across fifty. Nobody leaks for the money; they leak because three extra grams felt like insurance against a weak throw. It is not insurance. Past the wax’s stated maximum the extra oil is not held in the matrix, and the first thing it does is look for a route out (why more fragrance oil does not always mean stronger candles).

Fixing the pour: five steps that stop leaks at the bench

Two of those five steps cost nothing and two cost a thermometer. The fifth costs patience, which in a festive month is the most expensive item on the list — and the reason leaks cluster in October. If the calendar is tight, cut the range, not the cure.

Rescuing a batch that is already leaking

Start by separating the batch into three piles: jars that are dry on the bench and dry in the warm leg, jars that are dry cold and wet warm, and jars that are wet at room temperature. The first pile can go out. The second pile is stock you can sell locally and must not courier in summer. The third pile is raw material.

Remelt and dilute: the arithmetic

If the audit says the load is the problem, dilution fixes it exactly, because you are adding wax to a known quantity of oil. For every 1 kg of wax in a 12% batch, add 200 g of fresh wax to land at 10%: the oil stays at 120 g and the wax becomes 1.2 kg. The general rule is the same shape — new wax = old wax × (old load ÷ target load − 1). At ₹507.40/kg that 200 g is ₹101.48. For a fifty-jar batch of 150 g candles, 7.5 kg of wax needs ₹761.10 of dilution wax, and a ₹2,537.00 5 kg pack covers it with wax to spare.

Remelting costs you the wicks and the labels, not the oil. Scrape the wax out, remelt to the grade’s stated working temperature, add the dilution wax, stir, and re-pour into cleaned jars — Isopropyl Alcohol Spray at ₹383.00 takes the oil film off glass so new wax adheres to a clean wall. You will get more candles than you started with, so order jars and wicks accordingly: 7.5 kg of wax at a 150 g fill is fifty jars, and 9 kg is sixty. CSI’s guide to re-melting a failed candle walks through the handling.

Downgrade to wax melts — with one honest caveat

A wickless melt in a warmer has no flame, no jar wall to grip and nothing under it to stain, so a soft, heavily scented block is far less of a liability than the same wax in a lidded jar. Wax Melts Containers are ₹29.50 each in the pack of 10 (₹295.00). The caveat is the one nobody says out loud: the wax’s stated maximum does not change because the format did. Several CSI oil pages quote a higher range for melts than for candles — Lavender gives 10–15% for melts against 6–10% for candles — but if you are above the wax page maximum, dilute first and pour melts second.

When to write it off

Write off any jar where oil has reached the label adhesive, any tin that has passed oil through the seam, and anything you cannot trace to a batch card. The materials in a 150 g candle are ₹213.23; a refund, a replacement and a public review cost several times that. Sell the honest version of the story to your customers if some have already gone out — Indian candle buyers are forgiving about a replacement and unforgiving about a second identical parcel.

Packing and shipping soy candles so they arrive dry

Cure before you pack. A candle that goes into a carton at 48 hours is still settling; one that goes in at 14 days has done most of its moving. Cap the jar only when the surface is dry, wipe the rim and the outside of the glass, and check the base with a fingertip before the label goes on — a label applied over a film of oil will lift within a week whatever the courier does.

Then pack for heat and for orientation. Upright, every jar individually wrapped so a leak from one cannot mark its neighbours, void fill that holds the jar off the carton wall, and the label positioned above the fill line rather than across the base seam. Tins deserve an extra layer: put the candle in a sealed bag inside the carton so that a seam leak stains a bag and not the box. None of this stops a leak; it contains one and keeps the rest of the order saleable.

Two real reviews above are relevant here and neither is entirely flattering. One verified tin buyer likes the material and the mandala design but reports that some tins arrived a little dented and that the lid is hard to open once a finished candle is inside — a dent at the rim is exactly where a snug lid stops sitting flat. Another verified customer of Premium Coconut Soy Wax calls the products premium quality and says the delivery charges are high enough to restrict how much she buys. We do not publish delivery times or shipping costs on these pages, so the practical answer is to consolidate: order wax, jars, wicks and packaging together on one consignment and ask for bulk pricing on WhatsApp rather than paying for four separate parcels.

Finally, decide what you will not ship. A formula that passed a bench test but failed the warm leg is a shop-counter and local-delivery product until you have fixed it. There is no shame in a "workshop only" line; there is real damage in a Diwali order that arrives with four dry candles and two wet ones, because the customer will not remember which was which — only that your brand leaked.

Frequently asked questions