Headspace Math: Preventing Film Burst When Gel Expands on Freezing

A pouch filled to the shoulder with gel looks like good value and bursts the first time it freezes. The physics is not subtle — water expands about 9% on freezing — but the arithmetic that turns 9% into a fill weight gets skipped on more lines than you would expect. The result is a run of packs that pass a water leak test at filling and split along the gusset at -18°C.

Water expands 9%, and the pouch does not

Ice occupies 1.09 times the volume of the water it came from. A gel pack at 0.8% SAP is roughly 91% water by volume, so the whole pack grows about 8.2% when it freezes solid. The film does not stretch 8%: a 90 micron coex PE will give 3–5% elastically before the seal starts taking load, and a PET laminate gives almost nothing. The remaining 3–5% has to be headspace, or the pressure goes into the seam. Burst pressure for a well-made 500 g pouch is 1.8–2.2 bar; the pressure from constrained ice expansion in a rigid pouch can exceed 3 bar at the corners.

Working out the fill weight

Take the pouch’s usable volume — the volume below the seal, minus the gusset dead space if the pack freezes flat. A nominal 500 ml pouch usually holds 540 ml to the seal line. Fill it to 92% of that, 497 g, and you have 43 ml of headspace, about 8.6%, which covers expansion with a little margin. Fill it to 96%, 518 g, and you have 22 ml against the 41 ml you need. That pack bursts.

Two refinements. First, gel is not pure water; at 1.2% SAP the free water fraction is lower and expansion drops to about 7.8%, but the margin is thinner than people assume. Second, partial freezing matters: a pack that freezes from the outside in can trap liquid in the core and generate local pressure at a weak point, so headspace distribution is as important as total headspace. A pack that freezes flat with the headspace spread along the top edge behaves better than one where all the air collects in a corner.

Geometry decides where it bursts

Corners concentrate stress. A sharp 90° corner on a pillow pouch is where 80% of failures start, and a 6 mm radius on the seal die moves the failure point to the seal face, which fails at a higher pressure and usually weeps rather than splits. Gussets help only if the gusset is deep enough to open; a 20 mm gusset on a 500 g pack gives 15 ml of extra volume and a fold line that acts as a hinge. Stand-up pouches with a shaped bottom hold expansion better than flat pillow packs but cost 30–40% more in film.

Fill temperature also matters. Filling at 30°C and sealing immediately traps vapour that condenses as the pack cools, pulling a partial vacuum that makes the pack look underfilled and then over-pressurises on freezing. Let the gel settle to 20–25°C, or fill with a slight top-off after a 10-minute rest.

Testing it without a freezer full of scrap. Freeze ten packs at -25°C for 48 hours, flat and stacked three high with a 5 kg weight on top — stacking load is what turns a marginal headspace into a failure. Then thaw at 30°C and inspect every seal with a 0.5 bar pressure test or a simple squeeze under water. Record the fill weight of any failure; if failures cluster above 94% of nominal volume, your fill target is the problem, not the film.

One caveat: more headspace is not free. Above 12%, packs slosh, the gel does not contact the full face, and cold delivery drops. Stay in the 8–10% band, and specify fill weight by mass with a checkweigher rather than by eye. Our SAP for ice packs is dosed at 0.6–1.2%, and the lower end of that range leaves you a slightly wider headspace margin at the same fill volume.