Aggregate Moisture Correction (Free Water)

A mix design is written on the assumption that the aggregates arrive in a condition that essentially never occurs in a stockpile: saturated surface-dry, meaning every pore inside the stone is full of water and the outside is bone dry. In that state the aggregate neither gives water to the mix nor takes any, and the batch water is the mix water. It is a bookkeeping fiction, and a useful one, because it gives the arithmetic a fixed reference. Real aggregate is either wetter than SSD, carrying free surface water into the mix, or drier, and about to drink some out of the paste.

The correction has a wrinkle in it that catches people, and it is worth walking through slowly. Moisture content and absorption are both quoted against the oven-dry mass. The scale weighs the aggregate wet. So a tonne of sand at six percent moisture is not a tonne of sand plus sixty kilograms of water — it is 943 kilograms of sand plus 57 kilograms of water, because the six percent was six percent of the 943. Multiplying the stock mass straight by the moisture percentage overstates the water by a factor of one plus the moisture content: a few percent, small enough to look right on a printout, large enough to be three kilograms of water in every tonne of sand, every load, all season.

That is what the division by $1 + MC$ is doing. Work in oven-dry mass and everything is consistent: the dry aggregate is $m_{st}/(1+MC)$, and the free water it carries is that dry mass times the difference between what it holds and what it absorbs.

The correction runs in two directions at once, and doing only half of it is the second commonest version of this mistake. The water has to be adjusted, yes — but so does the aggregate batch weight. The mix design calls for a certain mass of aggregate at SSD; if you weigh out that number in wet stock you are short of actual aggregate by the mass of the water that came with it. Adjust one and not the other and the mix is wrong in a way that is hard to see: the water-cement ratio drifts, the paste volume drifts, and the yield drifts with them.

Note also what happens when the sign flips. A stockpile that has been sitting in the sun through a dry week can fall below its own absorption, and then the aggregate takes water out of the paste after mixing — sometimes over quite a long period, since a stone absorbs slowly. The load arrives stiff, someone adds water on site to make it placeable, and the water goes in uncounted and unrecorded. The concrete ends up at a water-cement ratio nobody designed and nobody knows. This is a real and routine failure mode in hot weather, and the correction on this page is what prevents it.

Two practical notes. First, sand dominates. Fine aggregate holds far more surface water than coarse aggregate — a fine sand at six or eight percent is unremarkable, where coarse stone rarely carries more than one or two — so a plant that measures only one thing should measure the sand. Second, absorption and moisture content are different kinds of quantity and deserve different treatment. Absorption is a genuine property of the rock: it belongs on the supplier's test report, it does not change from day to day, and a dense trap rock and a porous sandstone sit in quite different places. Moisture content is a weather report. It changes with rain, with sun, with depth into the pile, and within a single shift. Absorption gets looked up; moisture gets measured, and gets measured often.

Lightweight aggregate is the case where this arithmetic stops being trustworthy. It absorbs a great deal, it absorbs slowly, and it keeps absorbing after mixing — so there is no stable SSD condition to correct against. Lightweight aggregate is pre-wetted rather than corrected for, and the practice for it belongs in the specification, not on this page.