Four Grams of Sugar and the Pressure of a Car Tire

The 500 mL swing-top on the counter will, by Thursday, be holding back roughly the pressure in a car tire. That is the part nobody warned me about, because last month I couldn’t keep a culture alive long enough to have a pressure problem. Chloramine sank Diane’s disc in eight seconds and the whole first attempt was a survival story. A carbon-block filter on the kitchen tap fixed that in a single pass — chloramine is one of the few things activated carbon reliably strips that boiling won’t. The batch that came off the filtered water fermented cleanly. So now I have the opposite engineering problem: containing what a healthy culture produces.

Second fermentation is the interesting bit, and it is genuinely a pressure-vessel exercise. You bottle the tea while it still holds a little unfermented sugar, seal it, and let the yeast keep working in a closed volume. The alcohol stays dissolved. The CO₂ has nowhere to go, so it pressurizes the headspace and forces itself into solution. That is carbonation — not something you add, something you trap.

The numbers are the useful part. Fermentation runs the stoichiometry of

C6H12O6  →  2 C2H5OH  +  2 CO2
 180 g          92 g       88 g

so a gram of sugar yields about 0.49 g of CO₂. One “volume” of carbonation means one litre of CO₂ gas (at 0 °C, 1 atm) dissolved per litre of liquid, which works out to 1.96 g/L. Divide it through and you get a clean rule of thumb:

Roughly 4 grams of residual sugar per litre produces one volume of carbonation.

A pleasantly fizzy kombucha sits at 2–4 volumes — the same range as beer — which at room temperature corresponds to about 200–400 kPa (30–60 psi) of gauge pressure. A decent glass swing-top is comfortable to maybe 2–3 atmospheres. You can see how the margin gets thin, and why “kombucha bottle bombs” are a documented emergency-room entry and not a folk tale.

Here is the honest complication. If four grams of sugar per litre buys a full volume of fizz, and a hydrometer resolves about 2.6 g/L per gravity point (0.001), then an entire volume of carbonation is barely 1.5 points of specific-gravity drop. That is at the edge of what a cheap hydrometer can read and well past what my tongue can. Last month I complained that kombucha gave me no legible signal the way the sourdough jar did when I pointed a CO₂ sensor at it. I was half right. The gravity signal exists; it is just too small to steer by directly.

So you cheat with mechanics. Fill one PET bottle from the same batch and treat it as a pressure gauge you can squeeze — a “pet rock.” It softens when slack and goes rock-hard at pressure, and it fails by bulging rather than shattering. My schedule:

Bottle at ~1.008 SG, still faintly sweet, leaving a few g/L of fuel.
Fill one PET “gauge” bottle plus the glass ones.
Sit at 20–22 °C for two to four days.
Squeeze the PET daily; when it won’t dent, you’re at pressure.
Refrigerate to stop the yeast — and this matters — because cold liquid holds far more CO₂ in solution (Henry’s law), so a chilled bottle vents gently while a warm one erupts.
Open cold, over the sink, with your face out of the line of fire.

The other thing I got wrong last time was reading the culture itself. A healthy new pellicle forms flat on the surface, translucent to cream, sometimes bubbled — it is bacterial cellulose, and Komagataeibacter xylinus builds sheets pure enough that people tan them into leather. Kahm yeast is a wrinkled, dry, crinkled film: ugly, harmless, skim it. Mould is the disqualifier — raised, fuzzy, three-dimensional circles in blue-green-black, exactly like bread mould. Flat and slippery is fine; fuzzy and dry means the whole jar goes in the compost. I could not have told you that a month ago, and it is the single most useful thing I now know.

The pressure signal was there the whole time. I just needed a culture healthy enough to generate it, and a bottle I could squeeze to read it.