Still Liquid at Minus 8: Supercooling and the Packs That Never Really Froze

A shipper pulled twelve packs from a minus 22 freezer after 30 hours, found them rock hard, and loaded them. Two hours into the run, a handler dropped a crate and four of the packs turned to slush in his hands, releasing a wave of latent heat that pushed the box interior up by 3 degrees in eleven minutes. The packs had never crystallised. They were supercooled liquid sitting at minus 6, and the impact was the nucleation event that started the freeze.

Supercooling is a real and under-diagnosed failure in gel pack production and cold-chain packing, and it looks exactly like success until the moment it does not. This article explains why water in a gel pack can sit below its freezing point without freezing, how to detect it, and the four changes that stop it happening on your line.

Why a Water Gel Can Sit Below Zero Without Freezing

Freezing needs a nucleation site: a surface, a particle, or a crystal template that lets water molecules organise. In pure water in a smooth container, nucleation may not begin until minus 6 or even minus 12. In a gel pack, the SAP particles and the film surface normally provide plenty of sites, which is why most packs freeze at or just below 0.

But formulation choices can suppress nucleation. High levels of xanthan or guar raise viscosity and slow molecular diffusion to the growing crystal front. Surfactants and antifreeze-type additives depress the freezing point outright. And a very smooth, low-surface-energy film gives the water nothing to nucleate against. Combine a heavily thickened gel with a glossy film and a freezer that cools quickly, and you get packs that are cold, firm to the touch, and completely unfrozen.

The Tilt Test That Takes Ten Seconds

Take a pack from the freezer and tilt it 30 degrees. A properly frozen pack does not move. A supercooled pack flows like thick syrup, and the flow often triggers nucleation on the spot, which you will see as the gel turning opaque and stiffening. If a pack moves when tilted, it is not frozen regardless of what the surface feels like.

For a more definitive check, insert a probe into the geometric centre of a sacrificial pack. A frozen pack reads below minus 15 and stays there; a supercooled pack reads minus 4 to minus 8 and starts climbing within a minute of the probe going in, because the probe itself nucleates the gel.

What Happens on the Lane

A supercooled pack has not released its latent heat, so it still contains about 320 kilojoules per kilo of energy that will be released later, inside your customer’s box. When nucleation happens during handling, the pack warms to 0 degrees and stays there while it freezes, which is the worst possible behaviour: the pack absorbs heat from the payload instead of delivering cold for the first hour.

The effect is not subtle. In our tests, a box packed with four supercooled 1 kg packs reached 8 degrees at the payload in 9 hours. The same box with the same packs properly frozen lasted 26 hours. The packs looked identical at loading.

Four Changes, and Why Mild Weather Is Worse

First, freeze colder and longer than you think. A freezer set to minus 18 with poor air movement may never take a 30 mm pack below minus 10 at the core in a single night. Set the room to minus 25, allow 3 m/s air across the crates, and hold for a full 16 hours for a 1 kg pack rather than the 8 hours many plants allow.

Second, seed the gel. Adding 0.05 percent of a fine mineral powder, or simply leaving 0.5 percent of unmelted gel from the previous batch as a seed, gives nucleation sites throughout the pack. Some pack makers deliberately include a small rough-textured insert for this reason.

Third, cut the thickener. If a formulation needs 0.3 percent xanthan to stop separation, the SAP is not doing its job, and the thickener is both hiding the problem and suppressing nucleation. Fixing the SAP dose solves both.

Fourth, verify per batch. Two packs per pallet, tilt-tested at the freezer door, recorded on the load sheet. It costs twenty seconds and catches the failure before the truck leaves.

Counter-intuitively, supercooling incidents peak in mild weather, when freezers are opened more often and packs spend time in staging areas at 8 to 12 degrees before loading. A pack that has partially warmed and been returned to the freezer is the most likely to be supercooled, because it has already lost its original crystal structure and the second freeze has to start from scratch.

The safest rule is simple: never re-freeze a pack that has fully thawed and expect it to behave like a fresh one. If you must, extend the freeze time by 50 percent and tilt-test before loading.

Formulation is the upstream fix. A SAP with a consistent particle size distribution gives uniform nucleation throughout the pack, so the freeze front moves predictably from the surface inward instead of waiting for an accident.