Pellet quality, moisture and mill throughput: what the mill controls

All the water steam adds leaves again at the cooler, and often a little more. The feed then leaves drier than its formula: fewer kilos for the same dry matter bought. Mills call this loss shrink. It is not only moisture: dust, bags and scales are part of it too.

Published orders of magnitude are scarce. An industry survey in the United States, relayed by the trade press (Stark, Kansas State), reports an average total shrink of 0.81 %, from a gain of 1.09 % to a loss of 2.5 % depending on the mill. Trade-press articles give, without citing a source, 1 to 2 % of tonnage. None of these values replaces your own.

The right reference is therefore local: formula moisture on one side, moisture measured in the bag on the other, batch by batch and season by season. To reach 88 °C at the conditioner, the Kansas State guide works out about 6 % water added to a winter mash and about 4 % to a summer mash; a 2026 study (Renner et al.) looks at the combined effect of ambient temperature and conditioning temperature on the moisture added. The same setting does not give the same result all year.

In pigs, three Kansas State trials summarised by the Irish research body Teagasc (Badger et al., 2022) set feeds with 9.6 to 15.5 % fines against feeds with 83.6 to 90.4 %: feed conversion worsens by 5.9 %, 8.6 % and 6.4 %. A pellet with a low fines level still beats mash, by 4.1 to 6.7 %. Pelleting raises feed efficiency, provided the pellet reaches the feeder whole.

In broilers, results vary. A classic study (McKinney and Teeter, 2004) concludes that at least 40 % pellets at the feeder are needed for weight and conversion to progress. A 2025 conference presentation (Minson et al., not peer reviewed) finds, on the contrary, a better conversion as the share of fines rises. A 2025 meta-analysis on pellet binders (Ridla et al.) reports that binders raise PDI, hardness and moisture without changing intake, weight or conversion.

A better PDI therefore does not always turn into a gain on the farm. It is a tool for setting the process, not a measure of what the animal actually eats: hence the value of relating PDI to the fines recorded on the farm.

Adding water at the mixer, before the conditioner, binds particles by capillary action and has time to penetrate the particle, which steam, in 30 to 45 seconds, only partly achieves. On corn-based feeds, published trials range from 1 to 5 % added water: 1 and 2 % (Hott et al., 2008), 2 and 4 % (Buchanan and Moritz, 2009), 2.5 and 5 % (Moritz et al., 2003), up to 3 % (Lundblad et al., 2009), 1.5 % (Knarr, 2026).

In most of them, pellet durability and press throughput rise with the water added and, in several, the relative electrical energy of the press falls. In Moritz et al. (2003), PDI goes from 75.6 % with no water to 76.9 % with 2.5 % water, then to 79.6 % with 5 % (values reported by Muramatsu et al., 2015).

The effect still depends on the plant. Hott et al. (2008) compare a commercial mill and a pilot mill, with 1 and 2 % of a mix of water and a mould inhibitor: PDI rises in the commercial mill, not in the pilot mill, and the authors conclude that the result depends on the manufacturing technique. In a commercial mill studied in 2025 (You et al.), water added as a processing aid is among the variables unfavourable to PDI.

Water also lubricates the die. Too much water reduces friction and can cancel the effect of a thicker die; above 18 % total moisture at the die, the press can choke (Kansas State, 2021). Added water is therefore set on the moisture of the mash and on what the steam supplies, not blindly.

Added water is also water available to moulds. The right indicator is not total moisture but water activity (aw): the share of that water that micro-organisms can use. FAO reference point: storage moulds grow from a water activity of 0.70. The full explanation is on the page mouldy feed in storage.

A point of moisture kept is therefore worth it only if it does not push the feed past the threshold where moulds start, in the bag, in the bin and in transit. A still-warm pellet loaded into a cold bin condenses water against its walls (Gerbitz, 2024); in hot, humid climates, every shelf life shortens. The places where water comes back are described in condensation in bins and trucks.

There is no universal shelf life for a pelleted feed. What shortens it: water activity, temperature, condensation and opened bags. Measuring water activity in the bag is the final test.

The target setting for Linkamold Ultra L, +1 to 2 %, is a quantity of water: the water added at the mixer with the product. It is not the moisture found in the bag. Between the two sit the conditioner, the press and above all the cooler, which removes part of that water, more or less depending on the plant and the season.

What published trials lead you to expect in the bag is more modest. In a 2024 study on two cattle-feed mills (Budak and Bilgeçli), an organic acid blend added at the mixer left 0.65 to 1.4 points more moisture at the cooler outlet than feed with no addition. That study concerns a commercial product that is not a Linker product, and one of its authors works for the company that sells it; the quantity of water added is not stated, and neither durability nor water activity was measured.

To our knowledge, no independent study has measured added water, a mould inhibitor, PDI, press energy and water activity in the bag together. The result on your line is therefore measured on your line: moisture and water activity in the bag, over several batches, without exceeding the formula moisture. A trial run with you is the way to do it.