Gel packs, bricks, and phase-change materials: choosing refrigerants for 2–8°C

Why 0°C is both the point and the problem

Water is a spectacular refrigerant. Melting ice absorbs a large amount of heat at a constant temperature, it costs almost nothing, and it recharges in any freezer. That is why EPS boxes with frozen water-based packs remain, as a 2026 Scientific Reports cold chain study puts it, “one of the most widely used cold chain packaging modes” (Feng, Scientific Reports, 2026).

The problem is arithmetic: 0°C is below 2°C. A freshly frozen pack sits well below 0°C — it comes out of the freezer at the freezer’s temperature — and until its surface begins melting it can drive anything it touches below the floor. Freezing is not the rare failure mode in refrigerated shipping; it is the common one. A 2017 literature review in Vaccine found exposure to temperatures below the recommended range during shipments in 38% of studies from higher-income countries (Hanson et al., 2017). Most of those shipments were packed by people trying to keep product cold. They succeeded too well. Winter multiplies the risk, which is why it gets its own post.

Engineered PCMs: moving the melt point, paying for it

An engineered PCM is a material whose phase-change temperature has been placed on purpose — for 2–8°C work, a few degrees above the freeze floor. The chemistries are documentable rather than proprietary magic: paraffins such as n-tetradecane, which PubChem lists with a melting point of roughly 5.5–5.9°C, and eutectic organic mixtures tuned to a target melt point. Published vaccine cold chain work has used a tetradecane–lauryl alcohol mixture and a decyl alcohol–lauric acid eutectic, each maintaining product at 2–8°C (Energy Reports review, 2022).

What you buy with the higher melt point: a pack that is “frozen” — fully charged — at refrigerator temperature, and that cannot pull the payload below its own melt point no matter how it is placed. The freeze-risk conversation largely disappears, and with it some of the conditioning burden. What you pay: several times the per-pack cost of water gel, generally less latent heat per gram than ice (so more pack mass for the same heat absorption), and a charging step that needs its own temperature-controlled space and more time than tossing packs in a freezer. PCM packs also need pack-state verification just as much as water packs do — a half-charged PCM brick looks identical to a charged one.

Pack states and seasonal mixing

Most real packouts are not one refrigerant. They are a configuration of pack states — the same physical packs, prepared to different temperatures, doing different jobs:

Season changes the mix, not just the count. Summer configurations lean on frozen packs plus a buffer between them and the product. Winter configurations cut frozen mass and add refrigerated or even room-temperature packs, because in January the box’s enemy is the −10°C trailer, not the porch. Shipping the summer configuration in winter is how product freezes; shipping the winter configuration in a heat wave is how it cooks. The configurations are different packouts, and each one needs its own chamber data.

Conditioning discipline: the step that actually prevents freezing

Conditioning is the practice of holding frozen water-based packs at room temperature until the surface begins to melt — until the ice core moves — before they go in the box. It puts the pack’s working surface at 0°C instead of at freezer temperature. This is not folklore: the WHO’s vaccine transport guidance exists specifically to minimize freeze exposure through the correct use of passive containers and their coolant packs (WHO Vaccine Management Handbook, WHO/IVB/15.03, 2015).

Mass ratios are folk guidance. The chamber test is the spec.

Ask around and you will hear ratios: so many pounds of refrigerant per cubic foot of payload space, so many bricks per box size. Treat these as what they are — starting points for a prototype, not evidence. Published ratios cannot know your wall material and thickness, your pack placement, your product’s own thermal mass, your lane’s ambient profile, or your season. Two boxes with identical refrigerant mass and different pack placement can produce different outcomes at the product probe.

The refrigerant does not have a temperature strategy. The packout does.

The defensible artifact is a chamber-tested packout specification: this box, these packs, in these states, in this arrangement, run against a seasonal ambient profile, holding 2–8°C at the product probe for the rated duration. That is the standard Meridian builds to — the M-48 and M-72 are third-party chamber-tested for their full 48- and 72-hour rated holds — and it is the standard you should demand from anyone, us included. If a configuration change matters enough to make (different pack count, different state mix, different season), it matters enough to re-test; that workflow is covered in validating a packout: OQ/PQ.