How to Find the Minimum Fragrance Load That Still Gives Strong Hot Throw
Share
What counts as "the panel can tell"? Three people who did not pour the candles score both on a 1 to 5 scale at the doorway, in two sessions with the rooms swapped. A step passes if its mean is no more than 0.5 below the reference and no more than one person scored it lower both times. Those thresholds are our judgement; write yours down before you start.
How much oil and money does it take? Five candles on day 0 (10, 9, 8, 7 and 6%) use 60 g of oil; a confirmation pair adds up to 28.5 g. A 100 g bottle covers every path. With Roasted Coffee the materials come to ₹1,524.94 for seven candles.
What does the minimum save? With Roasted Coffee at the 1 kg price, 7% instead of 10% saves ₹16.82 of oil per candle, ₹14.54 after the extra wax, and stretches 1 kg of oil from 66 candles to 95.
- Straight answer: how to find the minimum
- Why search downwards from the ceiling, not upwards
- Three checks that decide whether the test means anything
- The ladder: what to pour on day 0
- The blind panel: people, rooms, codes and scores
- The stop rule, and a worked scoring sheet
- What the test costs and what the minimum saves
- What to buy for a minimum-load test
- Fitting it around Diwali and Christmas
- Frequently asked questions
Why search downwards from the ceiling, not upwards
Many makers find their load by going up: pour at 6%, find it a bit quiet, try 8%, then 10%. The trouble is the question at each step. "Is 8% strong enough?" has no reference point except memory and hope, so the answer drifts with mood, room and how long you have been pouring. Makers who search upwards tend to stop at the top of the range because the top feels safest, and the top is the dearest place to stop.
A downward search fixes the reference first. You pour the strongest candle your materials allow, the lower of the two stated ceilings, and every lower candle is compared with that same candle. The question becomes "can a blind panel tell this one is weaker?" That is a comparison, not an opinion, and it has a natural stopping point: the first load the panel can reliably separate from the ceiling. Everything above that point is oil you are paying for without anyone noticing.
This post is about the method: the descent, the panel and the stop rule. Choosing between 6, 8 and 10% on a single oil's stated range, with a three-load test, is covered in the post on choosing 6%, 8% or 10% for soy. Why extra oil stops adding strength at all is argued in Why Doesn't Adding More Fragrance Oil Always Make a Candle Stronger?. Here we assume you already pour at or near the ceiling and suspect you are overpaying.
| Upward search (6 → 8 → 10%) | Downward search (ceiling → 8 → 7 → 6%) | |
|---|---|---|
| The question at each step | "Is this strong enough?" | "Can the panel tell this from the ceiling?" |
| Reference point | Memory of earlier candles | A candle of the same age, burning in the next room |
| Where makers tend to stop | At the top, because it feels safe | At the last load the panel could not separate |
| Typical bias | Upwards: more oil feels like insurance | Downwards, if the panel is lazy; the two-session rule guards against it |
| What you learn | That the top of the range works, which the page already implied | How much of the top of the range you can drop without anyone noticing |
The downward search has one bias of its own: a tired or polite panel says "same" too easily. The stop rule below uses two sessions, swapped rooms and a "lower in both sessions" test to guard against it. And it has one precondition, which is the subject of the next section.
Three checks that decide whether the test means anything
Before you pour a ladder, be sure the reference candle is a fair ceiling. Three checks, all free if you own a scale and a thermometer.
Check 1 — is the ceiling really the ceiling?
Two ceilings apply and the lower one governs: your wax page's stated maximum and your oil page's stated candle maximum. Luxury Soy Wax CSI 464 states up to 10%; Luxury Soy Wax Chunks and Premium Coconut Soy Wax CSI 468 state up to 13%. Most oils in this post state 10%, so 10% is the reference in any of those waxes. Coffee Vanilla Tobacco states 7–9%, so its reference is 9% even in a 13% wax. Where an oil page states no candle load at all (52 of CSI's 102 oil pages do not state one), work inside the wax ceiling and say so on your sheet. These are formulation limits, not regulatory limits, and the maximum is a limit, not a target.
Check 2 — does the reference already throw well?
This is the "is it actually the oil?" question, turned round. The descending test assumes your ceiling candle fills a room once it has a full melt pool at day 14. If your current candles are weak even at 10%, the problem is almost never that you need to find a lower load; it is wick, cure, process or room. Stop here and use the weak-hot-throw triage or I Used 8% Fragrance Oil but My Candle Still Smells Weak — What Should I Change? first. A ladder poured under a weak reference will "prove" that 6% is as good as 10%, because both are equally poor, and you will have saved money on candles nobody wants.
Check 3 — is the wick settled at the ceiling?
Hold the wick constant down the ladder; one variable at a time. Settle it first at the reference load: a full melt pool to the glass within roughly 3 to 4 hours on the first burn is a commonly cited rule of thumb, not CSI data, so measure your own jar. Lower loads can burn a little differently, a tendency that varies by oil, so log pool width and depth for every rung and note anything odd; a separate wick question belongs to Does Increasing Fragrance Load Mean I Need a Bigger Wick?.
The ladder: what to pour on day 0
For an oil with a 10% reference, pour five candles on day 0: 10% (the reference), 8%, 7%, 6%, and a 9% insurance candle that you judge only if 8% fails. Pouring them together means every comparison is between candles of the same age, the same wax lot and the same temperature log. Staging the rungs over several weeks saves a few grams of oil and wrecks the comparison.
| Candle | Load | Oil | Wax | Role | Roasted Coffee oil, 1 kg price |
|---|---|---|---|---|---|
| R | 10% | 15 g | 135 g | Reference, burned in every comparison | ₹56.07 |
| S | 9% | 13.5 g | 136.5 g | Insurance: judged only if 8% fails | ₹50.46 |
| T | 8% | 12 g | 138 g | First test step | ₹44.86 |
| U | 7% | 10.5 g | 139.5 g | Second test step | ₹39.25 |
| V | 6% | 9 g | 141 g | Third step; the page's stated floor | ₹33.64 |
| Day 0 total | 60 g | 690 g | Five candles |
Why a two-point first step (10 to 8) and then one-point steps? Because the first comparison is the most informative: if 8% cannot be told from 10%, the whole top fifth of your load was invisible, and you have already found most of the saving. The finer steps below 8% then look for the rest. If 8% fails, the 9% candle tells you whether any saving is available at all.
Oils with a narrower range
Shift the ladder to fit the page. Coffee Vanilla Tobacco states 7–9%, so its ladder is 9% (reference), 8% and 7%: three candles and 36.0 g of oil. An oil stating "up to 10%" with no lower figure, such as Jasmine or British Rose, gives you no floor; stopping at 6% there is our judgement, not the page's, so ask on WhatsApp if you want to go lower. An oil stating 8–10% (ten CSI pages do, among them Wild Berries) leaves only one step, from 10 to 8; the ladder is short, and so is the saving.
The blind panel: people, rooms, codes and scores
A panel does not need to be large; it needs to be blind and repeated. Three people is our working minimum: family, a neighbour, a customer who likes your candles. None of them should know the loads or the letters. You, the pourer, are the helper: you light, place and time the candles, and you do not score.
Two rooms, swapped
Comparing one candle on Monday with another on Tuesday tests Tuesday as much as the candle. Burn the pair at the same time in two closed rooms of similar floor area, fans and AC off, doors shut. Next evening, swap the rooms. If one room is simply more generous, the swap cancels it. Write both floor areas on the sheet; a 10 m² bedroom and a 20 m² hall are not "similar".
Scoring
Each panellist scores each room at the doorway at 60 and 120 minutes, 1 (barely there) to 5 (fills the room), with at least ten minutes outdoors between rooms so the nose recovers. They also answer one forced question per session: "Which room was stronger?" A panellist may say "same"; that counts as not scoring the test candle lower. Nose fatigue is real; do not ask anyone to judge a third pair in one evening.
| Field | What to write |
|---|---|
| Pair | R (reference) vs T, U, V or S, by letter only |
| Candle ages | Both on the same day; day 14 or later |
| Rooms | Room names and floor areas; which candle was in which room each session |
| Scores | Each panellist, each room, at 60 and 120 minutes, 1 to 5 |
| Forced choice | Each panellist: stronger room, or "same" |
| Burn notes | Pool width and depth at 2 hours, soot, flame height, anything odd |
| Verdict | Pass or fail against the rule written on day 0 |
A panel is not a laboratory, and CSI has no measured throw data for any oil; this is a way to make a small maker's judgement fairer, not a scientific measurement. For a panel that compares different oils rather than loads of one oil, see Best Fragrance Oils to Test When Your Soy Candles Have Weak Hot Throw; for comparing two oils head to head, How to Compare Two Fragrance Oils for Actual Candle Performance.
The stop rule, and a worked scoring sheet
Our rule has three parts, and a step must pass all three. The thresholds are editorial judgement; you may tighten them for a premium range, but you must fix them on day 0.
| Test | Pass if | Why |
|---|---|---|
| 1 · Mean score | The lower load's mean (all panellists, both times, both sessions) is no more than 0.5 below the reference's | Half a point on a 1 to 5 scale is inside the noise of a home panel |
| 2 · Consistency | No more than one of three panellists scored it lower than the reference in both sessions | One person with a sharp nose is information; two agreeing twice is a result |
| 3 · Clean burn | Full pool by the usual time, no soot, flame as steady as the reference | A lower load that burns differently is a different candle, not a cheaper one |
Reading the ladder. Judge 8% first. Pass: judge 7%. Pass: judge 6%, the page floor for Roasted Coffee and Lavender. The first fail ends the descent and the minimum is the step above it. If 8% fails, judge the 9% insurance candle; if that fails too, stay at the ceiling. It is a legitimate result: some oils do need their full load in some systems, and knowing that is worth the test.
| Pair | Reference mean | Test mean | Gap | Lower in both sessions | Burn | Verdict |
|---|---|---|---|---|---|---|
| R 10% vs T 8% | 4.3 | 4.2 | 0.1 | 0 of 3 | Clean | Pass: step down |
| R 10% vs U 7% | 4.3 | 3.9 | 0.4 | 1 of 3 | Clean | Pass: step down |
| R 10% vs V 6% | 4.2 | 3.4 | 0.8 | 3 of 3 | Clean | Fail: stop |
| Confirmation: 7% vs fresh 10% | 4.4 | 4.1 | 0.3 | 1 of 3 | Clean | Repeated: produce at 7% |
In this illustration the minimum is 7%. The 7% pass was not unanimous (one panellist marked it lower both times), which is exactly why the confirmation pair exists: a borderline pass that repeats is a result; one that does not repeat sends you back to 8%. Your own oil may stop at 8%, at 6%, or at the ceiling.
What the test costs and what the minimum saves
The plan below follows the path of the worked example: five candles on day 0, a stop at 7%, and a confirmation pair at 7% and 10%. Other paths use slightly less oil; none uses more than 88.5 g, so the 100 g bottle covers all of them.
| Item | Used | Value consumed |
|---|---|---|
| Roasted Coffee 100 g (₹390.00) | 85.5 g (60 g ladder + 25.5 g confirmation pair) | ₹333.45 |
| Luxury Soy Wax CSI 464 at ₹0.51 per gram (10 kg price) | 964.5 g | ₹489.39 |
| Clear Glass Jar with Golden Lid (150 gram), pack of 10 ₹944.00 | 7 jars | ₹660.80 |
| Eco Candle Wicks Thin (C1), pack of 10 ₹59.00 | 7 wicks | ₹41.30 |
| Total | 7 candles | ₹1,524.94 (₹217.85 per candle) |
New spend, if you own no jars or wicks: ₹1,393.00 (oil, a pack of jars, a pack of wicks); add ₹354.00 for the scale if you measure by spoon (₹1,747.00 in all). If you already own the wax, jars and wicks, the only new cost is the oil. All figures exclude labour, freight, packaging and GST. The same plan with Lavender uses ₹461.70 of oil at the 100 g price.
What the minimum is worth, per candle
Lower load means less oil and slightly more wax, because the candle still weighs 150 g. Count both. The table uses 1 kg oil prices, which is where the saving lands once you produce at the new load.
| Load | Oil per candle | Candles per 1 kg of oil | Roasted Coffee | Lavender | Coffee Vanilla Tobacco |
|---|---|---|---|---|---|
| 10% | 15 g | 66 | ₹56.07 | ₹78.31 | above its 9% page maximum |
| 9% | 13.5 g | 74 | ₹50.46 | — | ₹116.09 (its reference) |
| 8% | 12 g | 83 | ₹44.86 | ₹62.65 | ₹103.19 |
| 7% | 10.5 g | 95 | ₹39.25 | ₹54.82 | ₹90.29 (its floor) |
| 6% | 9 g | 111 | ₹33.64 | ₹46.99 | below its 7% page floor |
Worked example: Roasted Coffee at 7% instead of 10% saves ₹16.82 of oil per candle; the extra 4.5 g of CSI 464 costs ₹2.28, so the net material saving is ₹14.54 per candle. Stopping at 8% instead nets ₹9.69. For Lavender, 10% to 7% nets ₹21.21; for Coffee Vanilla Tobacco, 9% to 7% nets ₹24.27, the largest per-point saving of the three because its oil costs the most per gram. At 7%, 1 kg of oil makes 95 candles instead of 66, 43.9% more. How those per-candle figures add up across 100, 500 or 1,000 candles, and the risk side of the ledger, is the subject of How Much Money Can a Candle Business Save by Reducing Fragrance Load From 10% to 8%?; whether a dearer oil at a lower load can beat a cheaper one is in Can I Reduce Fragrance Load by Buying a Better-Performing Fragrance Oil?.
What to buy for a minimum-load test
One tool and three oils. The tool comes first because a minimum-load test is only as good as its grams. The oils are chosen for what their pages state, a candle load range with a floor, a notes structure and prices at every size on 25 September 2026, not for throw, which CSI has not measured. Candle counts assume 150 g at 8%, the series standard; at your minimum the counts rise.
| Size | Price | Per unit | What it covers |
|---|---|---|---|
| Candle Making Weighing Scale, single | ₹354.00 | ₹354.00 | every oil and wax weighing in the test and in production |
| Size | Price | Per gram | 150 g candles at 8% |
|---|---|---|---|
| 15 gm | ₹93.22 | ₹6.21 | 1 |
| 50 gm | ₹230.00 | ₹4.60 | 4 |
| 100 gm | ₹390.00 | ₹3.90 | 8 |
| 500 gm | ₹1,890.00 | ₹3.78 | 41 |
| 1 kg | ₹3,738.00 | ₹3.74 | 83 |
| Size | Price | Per gram | 150 g candles at 8% |
|---|---|---|---|
| 15 gm | ₹105.02 | ₹7.00 | 1 |
| 50 gm | ₹270.00 | ₹5.40 | 4 |
| 100 gm | ₹540.00 | ₹5.40 | 8 |
| 500 gm | ₹2,620.00 | ₹5.24 | 41 |
| 1 kg | ₹5,221.00 | ₹5.22 | 83 |
| Size | Price | Per gram | 150 g candles at 8% |
|---|---|---|---|
| 50g | ₹449.00 | ₹8.98 | 4 |
| 100g | ₹899.00 | ₹8.99 | 8 |
| 500g | ₹4,399.00 | ₹8.80 | 41 |
| 1kg | ₹8,599.00 | ₹8.60 | 83 |
Fitting it around Diwali and Christmas
The full method takes about five weeks: 14 days of cure, about a week of evening comparisons, and a confirmation pair that needs its own 14 days. From 25 September 2026, that runs past the point where Diwali stock (8 November 2026) must be poured and cured. Our honest advice: produce Diwali at the load you already trust, run the descent alongside it, and bring the new load in for Christmas (25 December 2026).
| Date | Step |
|---|---|
| 28 September 2026 | Day 0: pour the five-candle ladder (R, S, T, U, V) from one melt. |
| 10–11 October 2026 | First full-pool burns; pool log only. |
| 12–17 October 2026 | Day 14 onwards: descend, one comparison per two evenings. |
| 18 October 2026 | Pour the confirmation pair (winning load + fresh reference) from a new melt. |
| 1–2 November 2026 | Judge the confirmation pair at day 14; decide the production load. |
| 8 November 2026 | Diwali, stocked at your old load. |
| From 3 November 2026 | Christmas production at the proven load; 14 days' cure leaves weeks before 25 December 2026. |
Once the load is proven, bulk pricing changes the figures: at the 500 g price, Roasted Coffee costs ₹56.70 per candle at 10% and ₹39.69 at 7%, and 500 g makes 47 candles at 7%. The 1 kg is 4.2% cheaper per gram than the 100 g you tested with.
CandleMakingSuppliesIndia sells the oils, wax, jars, wicks and scale named here, and does not sell finished candles. This page shows you how to use less of our oil, and tells you not to buy a new one for the test. That is deliberate: a maker who knows their true load orders accurately and comes back.
Every price and per-gram figure was verified on CSI's live store on 25 September 2026; every size shown on the pick cards is in stock on that date. Loads, notes, cure and throw wording are quoted from product pages; where a page is silent (the scale's resolution, Lavender's cure time), we say so.
The descending method, the two-room swap, the 0.5-point and one-in-three thresholds, the 6% floor for oils with no stated floor and the worked scoring sheet are editorial judgement and illustration; CSI has no measured throw data for any oil. For bulk pricing, GST invoicing, and MSDS and IFRA documentation on request, message us on WhatsApp at +91 7397976926.
Frequently asked questions
- How Do I Know If I'm Using Too Little Fragrance Oil in My Candle?
- How Do I Know If I'm Using Too Much Fragrance Oil in My Candle?
- Why Doesn't Adding More Fragrance Oil Always Make a Candle Stronger?
- Should I Use 6%, 8% or 10% Fragrance Oil for Strong Soy Candles?
- Should Every Fragrance in My Candle Collection Use the Same Fragrance Load?
- Can I Reduce Fragrance Load by Buying a Better-Performing Fragrance Oil?
- How to Optimise Fragrance Load Before Buying Fragrance Oil in Bulk
- How Much Money Can a Candle Business Save by Reducing Fragrance Load From 10% to 8%?
- I Increased Fragrance Oil From 8% to 10% but Hot Throw Didn't Improve — Why?
- Best Fragrance Oils to Test When Your Soy Candles Have Weak Hot Throw
Customer reviews: The six reviews at the top of this page are genuine, published Judge.me reviews left by CSI customers on the product pages named on each card — Lavender Fragrance Oil, Luxury Soy Wax CSI 464 (Flakes Form) For Container Candles, Roasted Coffee Fragrance. They are general product reviews, not assessments of the troubleshooting advice on this page. Names, star ratings, dates and products are as recorded by Judge.me and wording is unedited. "Verified buyer" appears only on reviews Judge.me recorded as verified — Arminder Kaur, Akanksha Sharma, Priyanka Purohit, Akansha, Tina Raha, Thenirvaofficial. Reviews below five stars are included as published, not filtered out.
Figures verified 25 September 2026, CSI live pricing: Roasted Coffee 15 gm ₹93.22, 50 gm ₹230.00, 100 gm ₹390.00, 500 gm ₹1,890.00, 1 kg ₹3,738.00; Lavender 15 gm ₹105.02, 50 gm ₹270.00, 100 gm ₹540.00, 500 gm ₹2,620.00, 1 kg ₹5,221.00; Coffee Vanilla Tobacco 50g ₹449.00, 100g ₹899.00, 500g ₹4,399.00, 1kg ₹8,599.00; Candle Making Weighing Scale ₹354.00; Pen Thermometer ₹354.00; Luxury Soy Wax CSI 464 10 kg ₹5,074.00; Clear Glass Jar with Golden Lid (150 gram) pack of 10 ₹944.00; Eco Candle Wicks Thin (C1) pack of 10 ₹59.00. All sizes listed in stock. Per gram = price ÷ grams, two decimals.
Non-price figures: 150 g candle, load as a percentage of total weight (10% = 15 g oil + 135 g wax; 9% = 13.5 g; 8% = 12 g; 7% = 10.5 g; 6% = 9 g). Candles per 1 kg of oil are whole candles, floor(1000 ÷ grams per candle). Wax ceilings as stated on wax pages (CSI 464 up to 10%; Chunks and CSI 468 up to 13%), formulation limits rather than regulatory ones; oil ranges, notes, cure and throw wording quoted from oil pages. Net savings deduct the extra wax at the CSI 464 10 kg price. The panel thresholds, the melt-pool rule of thumb, the day-14 judging point, the 6% floor for oils without one and all scores in the worked example are editorial judgement or illustration. Costs exclude labour, freight, packaging and GST.