I Need a Candle Formula That Can Handle Indian Summer Storage Without Sweating — How Should I Select Wax and Fragrance Load Before Commercial Production?

Choosing the loads: inside every ceiling, on either basis

Three stated ceilings apply to the worked example: CSI 464 "Fragrance Load Up to 10%", CSI 468 "Fragrance Load Up to 13%", and the Chunks page's "up to 13% fragrance load capacity" (its FAQ: "13% of total wax weight"). The oil adds a fourth: Fallen Leaves "6-10% for candles". Under the two-ceilings rule, the lower of the wax's and the oil's maximum governs, so for this oil the ceiling is 10% in every wax. None of these pages states clearly whether its figure is a percentage of total weight or of wax, and the conservative reading (of wax) allows less oil: 10% of wax is 9.09% of total; 13% of wax would be 11.50%.

Two things follow. First, the grid tests every wax at the same two loads, so a difference between rows is a difference between waxes. Second, the higher stated ceilings of CSI 468 and the Chunks don't come into play with this oil; with an oil whose page allows more, such as one stating up to 13%, you could add a higher rung in those two waxes only, still within the lower of the two ceilings. How ceilings combine in general is the ceiling-lookup post; how the basis works is the load-basis hub.

The screening grid and its bill of materials

Each cell is a wax at a load: two candles for the cycling protocol and one room-temperature control. Two replicates per cycled cell is the minimum that tells a real result from a one-off pour fault; the control tells a wax result from a process fault. Six cells, 18 candles.

Costs are worked examples and exclude labour, freight, packaging and GST. One wick for all three waxes keeps the surface comparison fair, but it is a stand-in: different waxes commonly want different wicks, so the winning wax still needs its own wick bracket before production (the method is in the wick posts; start from the late-burn hub for the final-third checks).

Running the grid

Two records make the grid worth its cost. The first is the cycle itself: the dates, the hours in each phase and the readings you logged in the warm and cool places. A grid that "passed" a cycle whose warm phase never got warm has passed nothing, so if your logged readings fall short of the phase you set, extend the cycle rather than signing the result. The second is the process line for each wax: add and pour temperatures, stir time and cooling spot. When the grid picks a wax, that line becomes the production process, and a winner poured differently in production is a different formula.

From pour to decision is about three to four weeks: a week's cure, 14 days of cycling, a few burn sessions. Summer runs roughly April–June; to have a screened formula and a signed release sheet before April 2027, pour the grid by mid-January (judgement). If you pour on 14 March 2027, 169 days from 26 September 2026, the answer lands just as summer starts, too late for the first dispatches.

Reading the grid: patterns and what they mean

Ranked most common first as editorial judgement; CSI publishes no stability data for any wax–oil pair, and why oils hold differently in the same wax is 3230's subject.

Row 3 is where a 7% rung earns its place: if 8% has no demonstrated margin in any wax, add three candles at 7% in your preferred wax (₹661.93 of materials in CSI 464) and rerun, so that 8% can serve as the margin rung for a 7% production load. Row 4 sends you to another oil, or to a scent for the cooler months.

The selection rule, and cost per candle

The rule is deliberately conservative. A formula that passed your cycle at exactly its production load passed your cycle, in this year's weather, on candles you made carefully. Producing one point below the highest load that passed gives you room for a hotter shelf, a longer courier route and a less careful pour day. So: (1) eliminate any cell with a score of 1 or more on a cycled unit, and void any with a marked control; (2) a wax qualifies for production at 8% only if its 9% cell passed; (3) among qualifying waxes, run a blind throw test at 8%; (4) if two are indistinguishable, choose the cheaper per candle.

Per 100 candles, CSI 468 adds ₹1,788.48 of wax cost over CSI 464 at 10 kg prices. That is worth paying if CSI 468 is the only wax that qualifies, or if its throw clearly wins a blind test. It is not worth paying for a higher stated ceiling you won't use. Costs exclude labour, freight, packaging and GST.

After selection, the formula is fixed: wax, load, wick (once bracketed), jar, temperatures. It then goes through the summer release sheet before dispatch; shipping heat has its own test in 3229; and how a summer formula behaves in winter (frosting, wet spots) is 3289's trade-off.

The three waxes and the test oil

Four products: the three waxes the grid screens and the worked example's oil. Chosen on page statements, price and stock on 26 September 2026; CSI has no stability data for any wax–oil pair, so the grid ranks them, not us.

Frequently asked questions