How Do I Build an Approved Wick Matrix Showing Each Jar, Wax, Fragrance, Fragrance Load and Validated Wick Size?

An approved wick matrix is a register, not a sizing chart

Almost everyone who asks for an approved wick matrix is actually asking for two documents at once, and only one of them exists. The first is a chart that takes a jar diameter and hands back a wick size. No CSI wick page publishes one, no honest supplier publishes one, and the reason is simple: the same glass with a different fragrance at a different load in a different wax burns differently, so a diameter on its own cannot carry an answer. The second document is a register of combinations you have tested and approved, with the evidence attached. That one you can build, and this page is about building it. If you want the general reasoning that narrows the starting point before you test, our live wick guide by diameter does that job and this page will not repeat it.

The unit of the document is the row, and a row is not a wick. A row is one vessel, one wax grade, one fragrance, one load, one wick family and grade, one configuration, one spacing, one date, one evidence reference and one approver. Eleven of the twelve columns are part of the key that identifies the row. Exactly one column — the wick, with its configuration and spacing — is the result. Once you organise the sheet that way, everything else in this page follows mechanically: you know which columns you must copy exactly, you know which column you are allowed to conclude, and you know instantly when a row has stopped being true.

That framing also tells you what the matrix cannot do. It cannot be extended by inference. You may not read a row for a 7 cm jar in Premium Coconut Soy Wax CSI 468 at 8% and apply it to a 7 cm jar in Luxury Soy Wax CSI 464 at 8%, however close the two sound. You may decide, on evidence you record, that two rows can share one test programme — but that decision is itself a test, with its own candles, and the matrix is the place you write down that you made it. A matrix whose rows quietly inherit each other is a wish list.

The twelve columns, and why each one earns its place

Twelve columns sounds like a lot for a document whose whole output is a wick. It is not. Nine of them exist because a candle maker somewhere lost a week to their absence, and the other three exist because a customer asked a question nobody could answer. Here is the sheet, column by column, with a plain note on whether the column is part of the key.

Column 5 and 7: the load convention, and the fill reading you actually poured

Fragrance load is the single most commonly mis-recorded field in any candle document, because "8%" has two meanings and they produce different candles. On the CSI convention, and the one the product pages use, load is a percentage of wax weight: 200 g of wax at 8% carries 16 g of oil and 216 g goes into the jar. The other reading starts from the finished fill: 150 g of total fill at 8% of wax weight divides into 138.89 g of wax and 11.11 g of oil. The worked sheet on this page uses that second reading, because both of the named vessels publish a fill capacity in grams — 150 g and 300 g — so the capacity is the sensible starting point. At 300 g the same split is 277.78 g of wax and 22.22 g of oil. Write the percentage, the convention and the two gram figures in the row. Then nobody has to re-derive them, and nobody has to guess which reading the person before them used.

Column 6: the ceiling that belongs to the row, not to the range

Two published limits apply to every row, and the lower of the two governs. Luxury Soy Wax CSI 464 publishes a fragrance maximum of up to 10% of wax weight; Premium Coconut Soy Wax CSI 468 publishes up to 13%. All three oils in the worked sheet publish a soy candle range of 6–10% of total wax weight — Dark Vanilla Fragrance Oil; Lavender Fragrance Oil; Woody Oud Fragrance Oil. So in this particular set the oil pages govern every row and the ceiling column reads 10% twelve times over, including the six rows in the wax that would otherwise allow 13%. Putting that in the row rather than working it out each time is what stops a production supervisor pushing a load to 12% "because the wax allows it". These are the manufacturers' stated formulation maximums, and MSDS and IFRA documentation is available on request on WhatsApp.

Columns 3 and 4: why the lot reference sits in the key

A row is evidence about the material you actually burned. Swap in a new consignment of the same grade and you have a question, not an answer — and the only way to settle it is to know which consignment the row was earned on. No CSI product page carries a batch or lot number, so the reference you record is the one on your own invoice and packaging, in your own format: date received, pack size, a sequential code you assign. It does not matter that the code is yours. It matters that two rows with different codes can be told apart, and that a retained wick piece in a labelled bag can be matched back to the row that approved it.

A worked sheet: two vessels, two wax grades, three fragrances

A small range generates more rows than people expect. Two vessels, two wax grades, three fragrances and one load is twelve rows — and a range that modest is a normal Indian D2C candle line in its second year. Add a second load and it is twenty-four. Add a third vessel as well and it is thirty-six. Nothing in that arithmetic is avoidable by being clever about wicks, because the wick is not what multiplies.

Below is the sheet for the twelve-row case, with the money attached so the planning question is visible. One thing to be plain about before you read it: the wick column is left for you to fill. CSI publishes no burn times, melt-pool diameters, flame heights, consumption rates or soot measurements for any wick in the range, none of those figures exists anywhere in our data, and we have not burned these twelve candles. The structure of the sheet is what this page can give you honestly. The approvals are yours to earn and yours to sign.

Read the last column before you read any other. The twelve rows run from ₹224.66 of material to ₹440.69, a spread of ₹290.89 — more than the cheapest row costs in total. And the wick is ₹9.44 in every single row. In the cheapest row that is 4.2% of the materials; in the dearest, 1.83%. The column that does the least to your cost sheet is the column that decides whether the other three columns are wasted, which is the entire reason this document exists. Nobody builds an approved vessel matrix or an approved wax matrix. The wick gets one because it is the cheap part that invalidates the expensive parts.

The vessel column is worth a second look too. The 300 g rows cost roughly twice the 150 g rows because they hold roughly twice the fill, and the jar itself is a larger share of the dearer rows. That matters for how you sequence the programme: if you are going to run out of patience or cash, earn the rows for your 150 g line first, where three candles of evidence cost ₹673.98 at the cheapest combination, rather than the 300 g line where the same three candles are ₹1,546.65 at the dearest. Evidence is cheaper in small jars, and a row is a row.

What the document costs to earn, at three depths of evidence

A matrix is a spreadsheet, so it feels free, and then the first honest attempt to fill it stalls because nobody costed the candles. Cost them now, at the three depths of evidence people actually choose between, and the planning conversation becomes a real one.

Three numbers in that table are worth saying out loud. The sheet most people should build — three candles a row — is ₹12,645.53 of material, of which the wick content is ₹339.84, or 2.69%. The exhaustive version, where every row earns all five configurations from scratch, is ₹63,617.17. And the gap between those two is the reason serious makers spend their effort on deciding which rows may share evidence rather than on buying more wicks. Sharing is a separate decision with its own test, and the matrix is where you record that you took it — which rows were tested, which were approved by extension, and on what grounds.

There is a second, quieter cost that the table does not show, and it is the one that kills matrices: the cost of the rows you do not notice you need. Every new fragrance a customer asks for is four new rows in a range this size. Every "can you do it in the bigger jar" is six. If the sheet is honest, those requests arrive with a visible price attached, and some of them stop being free favours. That is not an argument against taking the work — it is an argument for quoting it properly. Our live guide on maintaining consistent quality as production grows sits next to this one.

Versioning: how a row dies, and how many die with it

The versioning rule is one sentence: a row is never edited, only superseded. When something in the key changes, the old row gets an end date and a reason, a new row gets a new ID, and both stay in the file forever. The document grows; it does not get rewritten. That is what lets you answer the only version question that ever matters in practice — not "what is approved now" but "what was approved on the day we poured the batch the customer is complaining about".

Give the file a version number that increments whenever a row is added or ended, and put the version and date in the header of the printed copy at the bench. A bench copy with no version is worse than no bench copy, because people trust it. The table below is the count that makes the discipline concrete: in a twelve-row sheet, here is exactly how much of the document each kind of change destroys.

Two of those rows deserve a note. The load change is the one people argue about: moving from 8% to 10% of wax weight in CSI 464 stays inside the 10% maximum the wax page publishes and inside the 6–10% soy range the three oil pages publish, so nothing about it is out of specification — and it is still twelve new rows, because the load is in the key. Being allowed is not the same as being approved. The wick supplier change is the other: an identical label on a wick from a different source is a name, not a specification, and it can differ in braid, treatment, coating, tab height and tab diameter behind that name. Twelve rows is the honest count there, and it is the strongest practical argument for keeping the wick column short.

Keeping it alive: one owner, one file, one monthly audit

The document fails for organisational reasons far more often than technical ones. Three rules keep it alive, and all three are about people rather than candles. One owner. One named person adds rows, ends rows and signs approvals, and nobody else writes in the file. One file. A single spreadsheet, with a printed copy at the bench carrying its version number and date, and no second copy anywhere. The moment there are two files there are two truths, and the bench will use whichever one is nearer. One monthly audit. Ten minutes, once a month, checking that every product you currently sell maps to exactly one live row.

That monthly audit is the whole maintenance programme, and it is worth describing exactly, because it is short. Take the list of products you are selling this month. For each one, find the live row. If there is no live row, that product is being poured on an approval that has expired or never existed, and it goes on a list to be fixed before the next production run. If there are two live rows, one of them should have been ended and the bench has been choosing. If there is a live row with no product, you are holding wick stock against something you no longer sell. Four outcomes, one pass through the list, ten minutes.

The purchasing read in that table is the one that turns a quality document into a buying list, and it is where CSI's range makes life easy in a way worth stating plainly. Column 8 across a whole live sheet usually resolves to a very short list — in our range, realistically Eco Candle Wicks in Thin (C1) at ₹5.90 a piece or Thick (C2) at ₹9.44; plus Cotton Braided Wicks Ready Made at ₹12.00 a piece for vessels the page puts above 8 cm. Two pre-tabbed cotton sizes, not a ladder of gauges, and a third family for large diameters. Three lines is a reorder list a small business can actually hold, and the per-piece rate is identical at every pack size, so the pack you choose — down to a 20-piece pack or up to the ₹4,720.00 500-piece pack — is a cash-flow decision, not a unit-price one.

Two honest notes on that table, because they change what you may write in column 8. The by-the-metre families arrive untabbed, and CSI sells no loose wick tabs: the only tabbed cut-to-length option is the Braided Wick Roll with 100 wick tabs, so a row whose wick column reads "cut length of Eco Braided Wicks" is a row your bench cannot execute unless the tabs are on the shelf too. And wooden wicks come in one width — there is no wider or thinner wooden wick in the range to step to, so a wooden row that fails is not fixed by changing the wooden wick. It is fixed by changing family, load or wax, and each of those is a different row.

What this document does not do

It carries no burn figures, and neither does this page. No burn hours, no melt-pool width in millimetres, no flame height, no consumption in grams per hour, no soot grade. CSI publishes none of those for any wick in the range and none exists in our data. Every threshold in your measurement sheet is a number you set and sign, and the matrix records the decision, not a reading we supplied. A supplier who hands you a target pool depth for a 7 cm soy candle at 8% has taken it from somewhere else.

It is not a diameter chart and cannot be turned into one. Twelve approved rows do not add up to a rule about 7 cm jars. If the pattern across your own rows starts to look like a rule, that is a hypothesis worth testing, and the test is candles. The only published diameter statements in CSI's wick range are that the ready-made braided wick is for candles exceeding 8 cm and that the loose 2 inch eco wick is for 2 inch diameters. Everything else is silent on diameter, and silence is not a hint.

It is not a safety certificate. An approved row says that a configuration met thresholds you set, on the candles you burned, in the conditions you recorded. It does not certify a candle as safe, and no supplier can do that for you from a catalogue. Keep the general rules in force in production as well as in testing — trim before each burn, burn within the stated time, never leave a candle unattended, and discontinue any configuration that overheats its vessel. A row in a spreadsheet does not override what a jar is doing in a 42°C room in Ahmedabad in May.

It does not transfer. Another maker's matrix is not evidence about your candles, even for an identical vessel and grade, because their key included their glass, their consignments, their pour temperature and their room. Read someone else's sheet for its structure and build your own rows. The same applies in the other direction: your matrix is a commercial asset precisely because it is not transferable, and it is the thing that makes a two-size wick range sufficient rather than limiting.

And an honest word about us, since this page is published by the supplier whose wicks fill column 8. One of the wick reviews at the top of this page is four-star and asks for better delivery timelines; another asks us to work on pricing. We publish them as written rather than selecting around them. On pricing, the fact we can state plainly is that our cotton wick rate does not reward you for ordering early or ordering deep: ₹5.90 and ₹9.44 a piece at twenty pieces or a thousand. On documentation, GST invoicing is standard and MSDS and IFRA papers come on request on WhatsApp — which matters here, because a matrix with lot references in it is a document an auditor or a corporate buyer will eventually ask to see alongside those papers.

Frequently asked questions