Fruit, vegetables, prepared meals, dairy and ingredient streams. The broadest category on this site and the one where "typical" values do the most damage, because a strawberry and a diced potato are not the same job.
Starting moisture across food runs from roughly 70% to 92%. On a 1,000 lb batch that is a 200 lb spread in the water the condenser has to trap.
High-moisture fruit at 88% gives up about 878 lb of water per 1,000 lb batch dried to 2%. A prepared meal at 72% gives up about 714 lb. Same pallet, a 164 lb difference in the actual work — and a cycle recipe developed on one is wrong for the other.
Sugar content matters separately: it lowers collapse temperature, which caps how much heat can go in during primary drying. High-sugar fruit is slow for a physical reason, not a machine reason.
It is the most common cause of a food cycle running long, and it is free to fix.
Vapour has to leave the product and reach the condenser. Doubling loading depth more than doubles the distance the last of that vapour has to travel through an increasingly dry, increasingly insulating layer above it. The bottom of a deep tray finishes long after the top.
Uniform, shallower loading finishes faster and more evenly than the deep tray that felt more efficient at loading time. This is why tray flatness and shelf spacing are specification items rather than details.
Slices, dices, purees, whole pieces, and finished meals.
Slices and dices present flat to the shelf, so conductive architecture serves them well. Whole pieces and irregular product cannot make that contact, which is where radiant HVT earns its place. Purees and liquids load in pans and behave more like the botanical work — collapse temperature usually sets the cycle.
All of it starts the same way: water load, then removal rate, then machine.