How Do I Test a New Glass Candle Jar for Heat Safety Before Ordering 500 Pieces?

Why 20 jars, and what 10 clean burns can and cannot prove

A first screen on three jars, as in the glass-suitability hub, answers "is this jar worth pursuing?". Before 500 pieces the question changes to "how confident am I that jars from this listing behave like this, including the ones I have not seen?". Twenty jars is our judgement for a small business: enough to burn ten in different conditions, keep references and still have a reserve for a retest, and a single pack of 20 on the in-stock CSI 150 gram listings. It is not a statistical standard, and it is not the full programme a brand runs before selling a new container; that is in the small-brand safety programme.

Be honest about what a clean result means. Suppose, as a worked example, that 1 jar in 20 in a batch had a hidden flaw that would crack it in use. If you burned 10 jars picked at random, the chance that all 10 came from the good jars is 59.9%. Burn 20 and it falls to 35.8%. If 1 jar in 10 were affected, 10 clean burns would still happen 34.9% of the time; if 1 in 50, 81.7%. The protocol is good at catching problems that most jars share, such as a jar that runs hot with your wick, a finish that changes, or a shape that brings the flame to the glass. It cannot prove that no jar in 500 will fail. That is why the protocol ends with an incoming check on the batch you receive, and why every candle still carries burn instructions.

Before the first pour: fix everything but the jar

The protocol tests the jar. If anything else in the candle is still moving, a failure could be the wick, the load or the pour, and you will not know which. Settle these first and write them on the log sheet.

The protocol, step by step

Checking glass temperature without touching it

Hot glass looks exactly like cold glass. Many makers "check" a jar by touching the rim or holding the jar, which is how people get burned, and it gives no number to compare. Use a non-contact infrared thermometer instead (CSI does not sell one). Point it at the same spots, from the same distance, in every session, near the end of the session when the jar is at its warmest: the rim; the side wall just below the melt pool line; the side wall at the level of the flame in the final third; and the base edge, without lifting or moving the jar.

Treat the readings as comparisons, not verdicts. Infrared readings on shiny or clear glass are approximate and depend on the device, the angle and the surface, and we do not give a temperature at which glass is "safe", because no such figure exists on any CSI listing and CSI publishes no thermal test data. What the readings do reliably is show you differences: one jar much hotter than its five sisters at the same spot, a reading that jumps between sessions, or a side wall that runs hotter on one side than the other (often a wick drifting toward that side). Each of those is a reason to stop and find out why. What a jar that gets very hot in its final burns usually indicates, and the fix order, is in the final-burn heat guide.

Pass, fail and retest rules

Decide the rules before the first pour, so a borderline result cannot be argued into a pass when the order is waiting. These are ours, as a planning framework for a small business, not an industry standard.

A retest on three reserve jars costs about ₹676.38 in jars, wax, oil and wicks at the prices in this guide. It is cheap enough that there is no reason to wave a single failure through.

The log sheet

The log is what turns 20 jars into evidence. It is also what you compare the delivered 500 with, and what you show a corporate buyer who asks how you tested. Keep one row per jar per session.

What to buy for the protocol

Everything below was in stock on 26 September 2026. The jar picks are candidates for the protocol, not jars we call safe; the tool picks make the results trustworthy. Oil and wax for the worked example are Jasmine and Luxury Soy Wax CSI 464, both in stock.

The cost against 500 pieces

The worked example: 150 g candles at 8% in the Clear 150 g jar, with CSI 464 and Jasmine. Twelve poured candles need 1,656 g of wax (138 g each) and 144 g of oil (12 g each). Two 1 kg packs of wax cover the wax with 344 g spare; a 100 g and a 50 g bottle cover the oil with 6 g spare, not enough for a thirteenth candle. Twelve Eco C2 wicks come from a pack of 20, leaving 8 for retests.

The protocol is 9.4% of the 500-jar bill and 4.0% of the full 500-candle materials bill; without the thermometer, if you already own one, it is ₹4,105.60. Across the 500 candles, materials come to ₹221.95 a candle. Put another way, if 1% of the 500 candles, five of them, had to be scrapped for a jar problem you could have found, the materials alone would be ₹1,109.76, before labour, refunds, replacement shipping and the customer's opinion of your brand. The infrared thermometer is extra and is not sold by CSI.

Honest pricing notes: our 150 gram jars are priced flat per jar across packs of 10 and 20, so there is no listed discount for 500; the larger packs do not cut the wax or wick price either (CSI 464 is ₹507.40 a kg in the 1 kg pack and in the 10 kg sack, and a C2 wick is ₹9.44 in 20s and in 500s); only the oil gets cheaper per gram, ₹3.92 in the 1 kg bottle against ₹4.30 in the 100 g. For 500 jars, ask on WhatsApp about stock depth, batch consistency and whether a quantity price is possible. Every figure here excludes labour, freight, packaging and GST.

Timing: Diwali, Christmas and the 500

Diwali falls on 8 November 2026, 43 days after this guide's pull date of 26 September 2026. Our worked-example protocol takes about five weeks from jars arriving to an order decision, before delivery of the 500 (CSI dispatches from Pune and does not publish delivery times; confirm on WhatsApp), pouring, curing, labelling and delivering to your customers. For a jar that has not started testing by the end of September, a 500-piece Diwali launch in it is not realistic without cutting the protocol, and cutting the final third is the one shortcut we would never take. Use a jar you have already tested for Diwali and bring the new jar in for Christmas, 25 December 2026 (90 days away), or Valentine's Day, 14 February 2027 (141 days).

Frequently asked questions