Gelato Draw Temperature: What -5 to -9 C Really Does
Table of contents
Draw temperature is the last number you control before physics takes over. It decides how much water is already ice, how small the crystals are, and how much damage the hardening cabinet still has to undo. Get it wrong and no amount of blast freezing will give you back the texture.
What draw temperature actually measures
Draw temperature, sometimes called extraction temperature, is the temperature of the product at the moment it leaves the batch or continuous freezer. It marks the end of dynamic freezing, the phase where the mix is frozen while being scraped and agitated, and the start of static freezing, or hardening, where the product is chilled without agitation.
That distinction is the entire reason the number matters. Dynamic freezing is the only stage where new ice crystals are being created and kept small by the scraper blades. Once the product leaves the machine, no new crystals form. The remaining water simply grows onto the crystals that already exist.
So draw temperature is not really a temperature target. It is a decision about how much of your ice you make under agitation, where crystals stay small, versus how much you make in a cabinet, where they only get bigger.
The number your machine can actually hit
Here is where most advice goes wrong by quoting a single figure. There isn't one.
For ice cream in a continuous or horizontal batch freezer, the standard reference figure from Goff and Hartel is a drawing temperature of -5 °C to -6 °C, at which point the product has the consistency of soft serve. That is the number the textbooks and most equipment manuals give.
Vertical batch machines are a different story. Their geometry and scraping action let them push colder, and for standard mixes the coldest practical extraction sits nearer -9 °C. Gelato drawn from a vertical machine commonly lands somewhere between those poles.
| Machine type | Typical draw | What you get |
|---|---|---|
| Continuous freezer | -5 to -6 °C | Soft serve consistency, high throughput |
| Horizontal batch | -5 to -6 °C | Conventional gelato draw |
| Vertical batch | down to about -9 °C | Denser, stiffer extraction |
The practical conclusion: find your machine's number rather than importing someone else's. A recipe drawn at -6 °C on a horizontal machine and at -9 °C on a vertical one is not the same product, and neither figure is wrong.
Why about half the water is still liquid
This is the fact that reframes everything. At the draw temperature of roughly -5 to -6 °C, only about half the water in the mix is frozen. The dissolved sugars and salts depress the freezing point, so the remaining water stays liquid in an increasingly concentrated syrup.
Quick reference. At draw, roughly half the water is ice. The rest freezes during hardening, onto crystals that already exist, because nothing is agitating it.
The mechanism behind that curve is freezing point depression, and the concentration effect it produces is covered in freeze concentration. The total freezing power of your sugar blend is what the PAC value expresses, and it is the single biggest lever on where your product sits on that curve at any given temperature.
Once the gelato leaves the machine, the remaining water has nowhere to go but onto existing crystals. No new nuclei form, because nucleation in this system needs the undercooling and agitation that only the scraped surface freezer provides. That is why hardening speed matters so much: slow hardening means a small number of crystals each growing large, which is exactly the mechanism behind coarse ice crystal texture.
Drawing too warm and drawing too cold
Both errors are real and they fail in opposite directions.
Drawing too warm leaves too much unfrozen water heading into the cabinet. More of your total ice then forms statically, on fewer crystals, producing a coarser scoop. Warm draw also tends to mean the mix spent less time under the blades, so overrun comes in low and the body reads thin and wet.
Drawing too cold sounds safer and is not. Pushing a batch freezer past its comfortable extraction point overworks the mix, and in high-fat bases the prolonged shear can partially churn the fat toward butter granules. The product also becomes genuinely hard to extract: it packs in the cylinder, comes out in chunks rather than a ribbon, and traps air pockets when you pan it. In a display case that same over-frozen product often reads as too hard to scoop later.
| Symptom | Likely draw error | Where to look first |
|---|---|---|
| Coarse, icy after a day | Too warm | Draw temp, hardening speed |
| Thin body, low overrun | Too warm | Draw temp, mix balance |
| Buttery specks, greasy mouthfeel | Too cold | Draw temp, fat content |
| Chunky extraction, air pockets | Too cold | Draw temp, cylinder load |
From draw to hardening to the case
Draw is one point on a chain, and the next link decides whether the draw was worth anything.
After extraction, product should go straight to a blast chiller. Hardening in the reference literature targets -18 °C, and ideally -25 to -30 °C; below roughly -25 °C the product is stable for long periods without meaningful crystal growth. Industrial lines blast at -30 to -40 °C for exactly this reason.
The variable you control here is not the cabinet temperature but the geometry. A deep pan hardens from the outside in and leaves the core sitting warm for hours, growing crystals the whole time. Shallow, level pans harden fast and uniformly.
Then the product moves to the display case, which is a different temperature again, chosen so the frozen fraction lands where a spade can work it. That target is covered in what temperature to serve gelato, and the physical state that governs long-term stability at those temperatures is the glass transition.
Skipping the blast step and moving straight from machine to case is the most common way shops undo a perfectly good draw. It is also a leading cause of freezer burn over a product's shelf life.
How to measure it properly
Three habits separate shops that know their draw temperature from shops that guess.
Probe the product, not the machine. The cylinder display reads the barrel or the evaporator, not the gelato. Use a calibrated digital probe pushed into the extruded ribbon as it comes out, not into the pan five minutes later.
Measure at the same moment every time. Temperature climbs quickly once the product is out. Take the reading mid-extraction, at the same point in every batch, or the numbers are not comparable.
Log it against the recipe. Draw temperature only becomes useful as a control when you can tie it to a specific mix. A base with different total solids or a different sugar blend will behave differently at the same temperature, and the log is what turns that into knowledge instead of folklore.
Once the number is reliable, it becomes a diagnostic. A batch that draws warmer than usual at the same recipe and the same time is telling you something about the machine, the ambient load, or the mix, and it is telling you before the customer finds out.
Related Concepts
Try these numbers in your batch
Free balancer · No signup wall · Watch PAC, POD, MSNF update live
