Losing it at scale

A product that was right at twenty liters and is not at two hundred, with nothing in the recipe having changed.

The batch that stopped working

Every recipe is implicitly a recipe plus a vessel, plus a temperature profile, plus a surface-to-volume ratio.

Scale changes all three while leaving the ingredient list identical, which is why the recipe looks innocent.

Fermentation does not scale linearly

Surface area to volume falls as a vessel gets bigger. For a kombucha, where the culture sits on the surface and much of the activity is there, that means a large batch ferments proportionally slower and differently.

Which produces a batch that is under-fermented at the usual time, with more residual sugar, which is exactly the condition that pushes both thresholds in the bottle.

Temperature gradients appear in a large vessel: warmer at the top, cooler at the base, so different parts ferment at different rates.

And the inoculation ratio has to scale with volume rather than staying a fixed amount.

Heat and cooling change

A large batch takes far longer to heat and far longer to cool.

Longer at temperature means more flavor loss and more color change, so a hot-filled product at scale tastes more cooked than the trial did.

And slow cooling is time spent in the range where organisms grow, which is a microbial risk that did not exist in a small batch that cooled in minutes.

Fixing it

Scale in steps rather than in one jump, and treat each step as a new product to be tested.

Measure through the process rather than at the end: Brix and pH through fermentation tell you the rate has changed before the batch is finished.

Adapt the process: more culture, longer ferment, agitation, a taller narrower vessel, active cooling.

And re-run the warm-storage test at the new scale, because the thresholds were validated on a batch you no longer make.

References