Blast Freezing Pan Depth: Why Shallow Hardens Better
Table of contents
Two pans of the same gelato, same blast cabinet, same hour. One comes out with a tight, dense body. The other is coarse and icy through the middle. Nothing separates them but depth, and the physics that punishes it is not linear.
The cabinet sets the air temperature. You set everything else.
Freezing time does not scale with depth. It scales worse.
The classical model for freezing time in food engineering is Plank's equation. For a slab cooled from both faces it takes this form:
t = (ρ · L / (T_f − T_a)) · (P·a/h + R·a²/k)
where a is the slab thickness, h is the surface heat transfer coefficient of the moving air, k is the thermal conductivity of the frozen product, and for an infinite slab the shape factors are P = 1/2 and R = 1/8. The same equation with different factors covers cylinders and spheres, which is why it has survived a century of textbooks.
Everything in front of the bracket is fixed by your recipe and your cabinet. Everything that you control sits inside it, and the bracket has two terms. The first is linear in depth: it is the resistance of getting heat off the surface. The second is quadratic in depth: it is the resistance of dragging heat out of the middle. Double the depth and the second term goes up fourfold.
Take representative values for a blast cabinet, an air-side coefficient of about 30 W/m²K and a frozen-product conductivity of about 1 W/m·K, and the bracket looks like this.
| Fill depth | Relative hardening time | Share from internal conduction |
|---|---|---|
| 30 mm | 1.00x | 18 % |
| 40 mm | 1.41x | 23 % |
| 50 mm | 1.87x | 27 % |
| 60 mm | 2.37x | 31 % |
| 80 mm | 3.48x | 38 % |
| 100 mm | 4.76x | 43 % |
Filling a pan to 100 mm instead of 30 mm does not take three times as long. It takes nearly five. And the deeper you go, the more of the total is conduction you cannot fix by buying a colder cabinet.
Gelato is a much worse conductor than the ice inside it
The intuition that trips people up is that frozen things conduct well. Pure ice does, at roughly 2.2 W/m·K. Gelato does not, and the reason is the thing you deliberately put there: air.
Overrun disperses millions of air cells through the matrix, and air is an excellent insulator. Cogné and colleagues, measuring and modelling the thermal properties of ice creams for the Journal of Food Engineering (2003, vol. 58, pp. 331 to 341), found that thermal conductivity for a given formulation depends on the density of the sample, which is to say on the air fraction, and on temperature. More air, slower heat transfer.
This produces a genuinely awkward result for anyone chasing volume. The higher your overrun, the better the insulation you have built into your own product, and the more the quadratic depth term hurts you. A high-overrun mix in a deep pan is close to the worst case the equation describes.
What those extra hours actually cost
Time in the cabinet is not neutral. It is time spent inside the temperature band where ice crystals grow.
At draw temperature, around -5 to -9 °C, a large fraction of the water in the mix is still liquid. As the product cools further, that water crystallises, and the freeze concentration of the remaining syrup rises. The question is not whether ice forms. It is whether it forms as a very large number of very small crystals, or a smaller number of large ones.
Fast cooling nucleates many crystals at once and gives none of them time to grow. Slow cooling favours growth over nucleation, and growth is what your tongue registers as coarse. Ice crystal size is the single most direct lever on perceived smoothness, and pan depth is one of the cheapest ways to move it.
The centre of a deep pan gets the worst of every world: it is the last region to leave the growth zone, it does so through the slowest gradient, and it sits furthest from the glass transition condition that eventually locks structure in place. That is why a deep pan is so often perfect at the edges and disappointing in the middle.
Quick reference. Fill blast pans to a maximum of 50 mm and aim for 40 mm when the mix is high in overrun or high in sugar. Spread the product level with a spatula before loading, because an uneven surface hardens as unevenly as it looks.
Figure 1: same pan, same air, two depths, and the conduction path that separates them.
Practical depths for a real gelateria
A standard Napoli pan measuring roughly 360 x 250 mm gives you a useful conversion. At 40 mm fill it holds about 3.6 litres, or roughly 3.1 kg of gelato at 30 % overrun. At 60 mm it holds 5.4 litres and about 4.6 kg. The deep pan carries about 50 % more product and takes roughly 70 % longer to harden.
| Fill depth | Volume per pan | Mass at 30 % overrun | Best used for |
|---|---|---|---|
| 30 mm | 2.7 L | 2.3 kg | Delicate bases, very high overrun, urgent turnaround |
| 40 mm | 3.6 L | 3.1 kg | The default for a busy lab |
| 50 mm | 4.5 L | 3.8 kg | Acceptable ceiling for most mixes |
| 60 mm and above | 5.4 L and up | 4.6 kg and up | Only if hardening time is genuinely not a constraint |
The trade is obvious once it is written down: more pans, more shelf slots, more handling, in exchange for texture. In practice most labs are cabinet-limited rather than pan-limited, and running two shallow loads beats one deep load on both quality and total throughput.
Airflow is the other half of the bracket
Depth controls the quadratic term. Air controls the linear one, through h. A blast cabinet delivers a high surface coefficient by moving cold air fast, and every obstruction you introduce steals it back.
Do not stack pans directly on top of each other, since a covered pan is cooled from one face instead of two and behaves like a slab of double the thickness. Leave the manufacturer's rail spacing rather than cramming trays into every slot. Do not cover pans with film during hardening; cover them afterwards, for storage. And do not load a warm cabinet, because a machine still recovering from the last batch is running at reduced capacity exactly when you need it most.
For anyone weighing whether the cabinet is worth it at all, the comparison in blast chiller versus home freezer makes the h term visible: same target temperature, wildly different air, wildly different results.
A depth test you can run this week
Take one batch and split it across three pans at 30, 50 and 80 mm. Load them together into the same cabinet, on separate rails, with the same airflow. Log the core temperature of each with a probe pushed into the geometric centre, and record the time each one reaches -18 °C.
You will get your own version of the table above, calibrated to your mix, your overrun and your machine. Then taste all three the next day, from the centre of each pan, and again a week later after they have sat in storage and been through normal service handling. The difference in the week-old samples is usually larger than the difference on day one, because a coarse start gives recrystallization a head start it never gives back.
Related Concepts
- Blast Chiller for Gelato covers cabinet selection, cycles and targets.
- Gelato Draw Temperature explains the state the product is in when it reaches the pan.
- Ice Crystal Size and Gelato Texture is the quality outcome depth is buying you.
- Gelato Overrun Explained covers the air that slows your hardening down.
- Heat Shock in Gelato is what undoes good hardening later.
- Walk-In Freezer for a Gelateria is where the pan goes next.
