BrauMo BrauMo ← Blog

Food safety · 8 min read · · Updated

Measuring alcohol in kombucha: which method gives you a number you can defend

Knowing the limit is the easy half. The hard half is producing a number at the limit that holds up when a customer or an inspector asks how you got it. Most of the ways a small brewery measures alcohol are not accurate enough at the value that matters.

Kombucha sits in an awkward place for alcohol measurement. Nearly every affordable method was designed for beer or wine, where the answer is somewhere between 4 and 14 percent and being off by a tenth does not matter. Your answer is around 0.3 to 0.8 percent, and a tenth is the difference between a soft drink and a licensed alcoholic product. This guide covers what each method actually does, which ones survive scrutiny, what to run in a small brewery, and how to prepare a sample so the result means anything.

Why this is harder than it looks

Three things make kombucha a difficult sample. It is acidic, and acids interfere with several indirect methods. It carries dissolved carbon dioxide, which throws off anything density-based until the sample is degassed. And it is alive: a sample sitting in a warm vial keeps fermenting, so a slow trip to the bench gives you a higher number than the tank held.

On top of that, the accuracy you need is unusual. A method with a plus or minus 0.2 percent error is perfectly respectable in a brewery and useless to you, because at a 0.5 percent limit that error spans the decision.

The methods

What KBI approves, and the instruments built for this

Kombucha Brewers International keeps a list of accepted ethanol testing methods, and it is the most useful page a producer can read on this subject. Its board recommendation for demonstrating compliance is headspace gas chromatography, with either flame ionisation or mass spectrometry detection. Enzymatic and gas chromatography methods for kombucha have gone through AOAC First Action status, which is what "validated for this product" looks like on paper.

The same list names in-house instruments it treats as acceptable, which is the part most producers are actually looking for.

Two instruments worth knowing are not on that list, so treat them as things on the market rather than as anything KBI has assessed. CDR KombuchaLab puts the enzymatic reaction into a benchtop unit sized for a production room rather than a laboratory, giving an alcohol result in about eleven minutes and reading sugars and pH on the same instrument. More general analysers sold into beer, wine, cider and kombucha, such as the ABV Technology unit, read from around a tenth of a percent upwards. If you buy either, the two-tier pattern below is what shows it agrees with a reference method.

Two cautions from that list are worth repeating. KBI treats hydrometers and refractometers as traditional in-house tools rather than as compliance methods, noting they are not finely attenuated enough to verify a limit, which is the same conclusion the arithmetic reaches. And a method carrying roughly a percent of variance can put you in violation while remaining scientifically respectable, because the variance is twice the limit you are measuring against.

What a small brewery should actually do

Two tiers, and the reason for two is cost.

  1. Screen in-house on every batch with an enzymatic assay or an analyser you have checked. This is your routine number, the one that goes on the batch record and tells you a batch is behaving.
  2. Verify externally at intervals with a lab running gas chromatography or distillation. Send a sample when you start a new recipe, when a process changes, when a batch reads close to the line, and on a regular cadence besides, quarterly for many producers. Keep the lab reports.

The pairing is what makes the in-house number credible. On its own it is your own instrument marking your own homework. Against a periodic external result that agrees with it, it becomes a screening method with demonstrated agreement, which is a much better thing to show an auditor.

Sample preparation, where most results are lost

When to measure, and the shelf-life question

The release measurement belongs at packaging, because that is the product a customer receives. For unpasteurised kombucha there is a second question that honest producers face: the product is still alive on the shelf, and alcohol can rise in a warm distribution chain.

If you sell raw and label the product as non-alcoholic, the defensible position is to know what happens over shelf life rather than to hope. That means testing held samples at intervals across the stated life, under storage that reflects reality rather than a laboratory ideal, and keeping the cold chain evidence alongside. If the numbers climb, the answers are process ones: finish the ferment further, reduce residual sugar, tighten the cold chain, shorten the stated life, or pasteurise. Retention samples exist exactly for this.

What the record has to contain

What an auditor asks

  1. What is the alcohol content of this lot, and how do you know?
  2. Which method produced that number?
  3. When did an independent lab last confirm your in-house method agrees?
  4. Does the figure hold to the end of shelf life, and what shows that?
  5. What happened the last time a batch came back over the line?

The third is the one that separates a screening method from a number you simply asserted, and it is the one most producers cannot evidence on the spot.

Frequently asked questions

Can I just use a hydrometer?
For watching a ferment, yes, and it is a useful habit. For the number you put on a specification or defend to an inspector, no. A hydrometer infers alcohol from sugar loss on the assumption of a yeast-only fermentation, and kombucha's bacteria and acids break that assumption. At around half a percent the inference is not reliable enough to carry the decision.
Do I need a lab for every batch?
Most producers cannot afford that and do not need it. The workable pattern is an in-house screen on every batch plus periodic external confirmation that your screen agrees with a reference method. Send a sample when a recipe or process changes, when a result sits close to the line, and on a regular cadence.
Should I test at packaging or at the end of shelf life?
At packaging for the release decision, because that is the batch you are letting go. Across shelf life as well if you sell unpasteurised and label it non-alcoholic, because the product keeps fermenting and the label has to stay true for as long as the product is on sale. Held retention samples are what make that possible.
A batch came back over the line. What now?
Treat it as a deviation and write it down: quarantine the lot rather than shipping it, confirm the result before acting on a single measurement, and decide what happens to the stock. Then look upstream, because a batch over the line is usually a process signal rather than a measurement accident: residual sugar, ferment length, conditioning temperature or cold chain. Record the investigation and the decision. A documented deviation with an action beside it is a functioning system; a batch that quietly shipped is the finding.
BrauMo keeps the alcohol result on the batch with its method, its sample point and the lot that shipped, alongside the acidification readings and the instrument records behind both, so the two control points are one story rather than two folders. Documentation-only: it records and reports, it never blocks a batch. See how the records are kept →

Related reading

See how BrauMo handles this: Take the product tour → · See pricing →