Freeze drying 101

Water has a trick: below a certain pressure it cannot exist as a liquid at all. Lower the pressure in a chamber far enough and ice turns straight to vapour without melting first — sublimation. A freeze dryer is a machine for doing that to food. It freezes the trays solid, pulls the chamber down to a vacuum a few hundred times thinner than the atmosphere, then gently warms the shelves so the ice sublimates out of the food and the water leaves as vapour.

That vapour has to go somewhere, or the pump would be trying to swallow it. It goes to the condenser: a very cold surface inside or behind the chamber that the vapour freezes back onto as frost. When the cycle ends, the condenser is defrosted and the water drains away. On home units the condenser is built into the chamber walls; on lab units it is a separate coil, and it is specified by how many litres of ice it can hold in 24 hours and how cold it runs. 24 of the 39 units here publish a condenser temperature.

Because the water leaves as ice, the food’s structure is left standing rather than collapsed by heat — which is why freeze-dried food rehydrates to something close to fresh, and dehydrated food does not.

Batch capacity is pounds of fresh food that fit on the trays for one run. It describes the machine. Across the 26 units here that publish it, the range runs from the Kolice FKD-05 at 4.4 lb to the Cuddon FD100LT at 220 lb; the median is 21.2 lb.

“Up to N pounds per year” is batch capacity multiplied by an assumed number of cycles, and the assumption is never printed. So this site prints it. Dividing the annual claim by the batch capacity gives the cycles the claim needs; multiplying by the maker’s own published cycle time gives the running hours. A year has 8,760 of them. Of the 9 units here that make an annual claim and publish a cycle range, 1 would need more hours than that at the long end of their own range.

The steepest assumption belongs to the Stay Fresh Large: its claim of up to 7,000 lb needs 255 cycles, or 5,091 h–10,182 h of running at its published 20–40 h cycle. The most conservative is the Blue Alpine XL Commercial, at 4,800 h–4,800 h.

None of this makes the annual figure false. It makes it a ceiling that assumes no time for defrosting, cleaning, loading or pump oil between runs. Read batch capacity as what the machine does; read the annual figure as what it could do if you never stopped.

The size-normalised price is dollars per pound of batch capacity. Among the 19 units with a published price and batch figure, the Stay Fresh Mega is the cheapest per pound at about $107/lb and the Kolice FKD-05 the most expensive at about $795/lb. Both are estimates from list price.

Home units are loaded on flat stainless trays that slide onto heated shelves. Tray count times tray width times tray depth gives the tray area, which is the real constraint: food is spread in a layer no more than about an inch deep so the ice can sublimate out of it, so what you can process per batch is a function of area far more than of chamber volume.

Pounds per tray — batch capacity divided by tray count — is the number to compare across brands, because a “7-tray” unit with small trays can hold less than a 5-tray unit with big ones. Both figures are shown on every unit’s page where the maker publishes tray dimensions, labelled as arithmetic.

Lab benchtop units mostly have no trays in this sense: samples go in flasks on a manifold, or on shelves inside a small chamber. They publish shelf area or chamber dimensions instead, and this site records those in the chamber field.

The vacuum pump runs for most of every cycle and is the part you will maintain. An oil-sealed rotary vane pump is cheap and pulls a deep vacuum fast, but water vapour from the food ends up in its oil; cloudy oil pulls a worse vacuum, so it has to be drained and filtered or replaced every few batches, and the pump vents a fine mist while it runs. An oil-free scroll or diaphragm pump has nothing to change and no mist, costs several hundred dollars more, and on some designs takes longer to reach full vacuum at the start of a run.

What ships in the box varies by maker and by generation, which is why the catalog records the pump each unit includes rather than the pumps a brand sells. Grouped by what is included: Oil (26), no pump (8), Oil-free (2).

Two different questions. Circuit is what the wall has to supply: small and medium home units run on a standard 110–120 V outlet, though several makers specify a dedicated 20 A circuit so a microwave on the same line does not trip it; large and extra-large units generally need 220–240 V, the same as an electric range. Of the units here that publish a voltage, 23 run on 110 V and 11 on 220 V.

Power draw is what you pay for. Watts times cycle hours, divided by a thousand, is kilowatt-hours per batch. Makers publish either a peak draw or a typical one, and a peak figure overstates the energy per batch because the machine spends much of the cycle below it; each unit’s page says which kind of figure it has. Among the 10 units with both a published draw and a cycle range, the estimate runs from 25.0 kWh ($4.24) per batch for the Blue Alpine Medium Select to 44.9 kWh ($7.63) for the Blue Alpine Large Elite, at $0.17/kWh, the 2025 US residential average.

A cycle has four phases: freezing the trays, pulling vacuum, the long drying phase where the shelves warm and the ice sublimates, and a final dry to pull the last few percent of moisture. Makers publish a range rather than a number because the drying phase depends on how much water is in the food and how thick it is spread. A tray of watery fruit takes far longer than a tray of cooked meat; a tray loaded two inches deep takes far longer than one loaded to three-quarters of an inch.

Published ranges here run from 12–20 h on the Blue Alpine Medium Select to 20–40 h on the Stay Fresh Large. Ambient temperature matters too — a hot garage makes the condenser work harder — and so does a pump pulling a poor vacuum on tired oil.

Smart sensing is the maker’s term for a controller that decides for itself when the food is dry, from chamber pressure or shelf temperature, instead of running a fixed timer. It ends cycles earlier on dry loads and later on wet ones. The catalog records whether each unit claims it.

The 39 units here split into 20 home, 12 lab benchtop and 7 commercial. Home units are sold in trays and pounds per batch, with a pump in the box and a controller that runs the whole cycle unattended. Lab benchtop lyophilizers are sold by condenser capacity and temperature, often without a pump, for drying samples in flasks — they rarely publish a food figure at all, and this site does not invent one. Commercial units publish batch capacity in kilograms, frequently need three-phase power, and are mostly sold on quote.

Size classes are the makers’ own tiers. Here they are: Benchtop (12), Medium (9), Commercial (7), Large (6), XL (3), Small (2). A “Medium” from one brand is not a “Medium” from another; read the batch figure.

Freeze-dried food leaves the machine at around one to three percent moisture, and it will take that moisture straight back from the air if you let it. The two things that spoil food are water and oxygen, so storage is about excluding both: a mylar bag (metallised polyester, effectively impermeable to water vapour and light) heat-sealed with an oxygen absorber (an iron-powder packet that scavenges the oxygen left in the bag) inside. Sized right, the absorber pulls the bag’s oxygen down to a fraction of a percent within a day.

Stored that way, in a cool dark place, freeze-dried food is routinely rated at 10 to 25 years by commercial producers. Fats are the limit — high-fat foods go rancid on a shorter clock regardless of the packaging. Jars with a vacuum-sealed lid work for food you will eat within months; they are not a long-term container, because glass lets light in and lids let a little air through.

Label every bag with contents, date and weight. This site’s whole look is borrowed from that label.