Refermentation in the bottle

The single mechanism behind both of this property's serious failures: the legal one and the physical one.

One cause, two failures

Fermentable sugar plus viable organisms plus time produces alcohol and carbon dioxide.

The alcohol is the 0.5 percent threshold. The gas is the pressure.

Which is convenient, in a way: control one and you have controlled both, and any evidence about one tells you about the other.

What starts it

Intentional live culture, in a kombucha or water kefir, which is the product working as designed and simply continuing.

Wild yeast, from fruit, from the air, from a hose. A juice or a sweet tea with no intended culture can still ferment, and it usually surprises the maker.

Culture carried over from a previous product on shared equipment, which is why the cleaning page insists on a full changeover between fermented and unfermented runs.

Why a stable product restarts

The frustrating case: a product that was stable in trials ferments in production.

Usually one of four things. A heat treatment that was not validated and did not fully kill. A filtration that passed organisms. Recontamination at filling. Or a batch with more residual sugar than the tested one, which gives whatever survived more to work with.

All four are process rather than recipe, which is why the answer is measurement rather than reformulation.

Stopping it

Remove the fuel: ferment drier, or sweeten with a non-fermentable sweetener.

Remove the organisms: validated heat treatment, or fine filtration, both of which end the live positioning.

Slow it: cold, throughout the chain, which is a control rather than a cure.

Add a hurdle: a preservative effective at your pH, which works and is a clean-label decision.

And prove it: hold retained bottles warm for the full shelf life and measure alcohol, pressure and Brix at the end. A product that has not moved on that test is one you can describe honestly.

References