Wine blending and the Pearson square: hit a target every time

Blending is where a wine is finally made — balancing components for alcohol, acidity, structure and aroma. Most of the numeric side comes down to two operations: working out what a blend will measure, and working out the proportions to hit a target. Both are simple, and both are worth getting exactly right before you move volume.

A blend is volume-weighted

For any measured property that mixes linearly — alcohol, sugar, titratable acidity, SO₂ — the blend value is the volume-weighted average of the components. Mix 1,500 L at 13.5% with 500 L at 12.0% and you get 2,000 L at (13.5×1500 + 12.0×500) ÷ 2000 = 13.1% vol. pH does not mix linearly (it is logarithmic), so blend on TA and measure the pH afterwards.

The Pearson square: proportions for a target

When you want a specific result, invert the problem. The Pearson square gives the ratio of two components to reach a target that lies between their values:

• Parts of A = target − value of B.
• Parts of B = value of A − target.
• Then normalise to fractions and multiply by your batch volume.

The target must lie between the two components. To reach 13.0% from a 13.5% and a 12.0% wine: parts A = 13.0 − 12.0 = 1.0; parts B = 13.5 − 13.0 = 0.5 → two parts A to one part B, i.e. 67% / 33%.

Always bench-trial

The maths gives you the starting ratio; the glass decides the wine. Aroma, texture and colour do not reduce to a single number, so build the calculated blend at bench scale, taste it, and adjust before committing tanks. Then re-analyse the real blend — small measurement errors compound at volume.

In GrapeFlow a blend is a real operation: the new lot inherits the exact volumes and origins of every component, so the composition is traceable straight back to the blocks — and forward to every bottle, for recall.