Dilution Water-Temperature Adjustment
A litre is not always a litre. Liquids swell as they warm and shrink as they cool, because their density changes with temperature — so a litre of spirit measured at 30 °C holds less alcohol than a litre of the same spirit measured at 20 °C, and a litre of warm water weighs less than a litre of cold water. The plain dilution tool assumes everything sits at 20 °C; this one lets the spirit and the water each be at their own temperature and works out the water you actually need at the temperature you measure it. The difference is small on a single bottle and worth having on a full batch.
A gram is a gram at any temperature, so the weight figure below is exact whatever the water temperature is — only the measured volume shifts. Weighing does not fix the spirit side, though: a litre of warm spirit still holds less alcohol than a litre of cold spirit, and that part of the correction only goes away if the spirit is weighed too, or left to reach 20 °C before measuring.
How the numbers are worked out
Everything runs through mass, because mass is the one quantity temperature cannot change. The spirit's mass comes from its volume times its density at its own temperature, looked up in the OIML R22 density tables; the share of that mass that is ethanol is fixed by the starting strength. The target strength then fixes the final mass — the ethanol mass divided by the target strength as a mass fraction — and the water needed is simply the final mass minus the spirit mass. That water mass is turned back into a volume using the density of water at the water's temperature, and the final volume is the final mass divided by the density of the finished spirit at 20 °C, which is the temperature a hydrometer or alcoholmeter reading refers to.
The correction pulls in two directions at once, which is what makes it worth a tool of its own. Warm spirit is thinner, so a litre of it carries less ethanol and needs less water: 1 L of 80% spirit at 30 °C proofed to 40% with 20 °C water needs 1.028 L of water, 10 mL less than the 1.039 L the plain 20 °C calculation gives, and the finished batch is 1.980 L at 20 °C rather than 2.000 L. Warm water pushes the other way, because the same weight of water takes up more room when it is warm: measuring that water at 30 °C instead of 20 °C adds about 3 mL to the volume. Cold spirit reverses the first effect — the same litre at 10 °C holds more ethanol and needs 10 mL more water than the 20 °C figure. At 20 °C for both liquids the two tools agree exactly, and the correction reads zero.
Two practical notes. First, the ABV you enter should be the spirit's true strength at 20 °C: if you gauged it warm with an alcoholmeter, correct that reading first (see the Temperature Correction tool), because a warm reading flatters the strength and the error would carry straight through. Second, the density tables run from 0 °C to 40 °C, and below 15 °C the values are a straight-line extension of the 15–20 °C slope rather than tabulated points — close enough for proofing (within about a millilitre per litre of water near 5 °C), but the tool refuses temperatures outside 0–40 °C rather than guessing further.
Ethanol–water densities at 15–40 °C come from the OIML R22 (1975) alcoholometric tables, as tabulated in the site's shared data module; the water column of those tables agrees with the Kell (1975) water-density values to the printed precision (0.99910 at 15 °C, 0.99820 at 20 °C, 0.99565 at 30 °C). The 20 °C answer is cross-checked against TTB Table 6 (27 CFR 30.66), which gives 1.040 L of water for 1 L of 80% spirit proofed to 40%.
See also: Dilution & Proofing, Dilution by Weight, Temperature Correction.