Carbon dioxide is one of the few ingredients in a carbonated drink that the consumer never sees on the label, yet it touches every bottle. The International Society of Beverage Technologists (ISBT) sets the quality guidelines that define what "beverage-grade" CO₂ means, and those guidelines are the reference the whole supply chain works to. From the producer purifying the gas to the bottler accepting a tanker at the gate. This article sets out what the ISBT CO₂ quality guidelines actually require, and why each limit sits where it does.
Where the guidelines came from
The ISBT guidelines exist because contamination in CO₂ has caused real problems. The most cited example is a benzene-in-CO₂ incident in the United Kingdom that prompted the industry to tighten its expectations and standardize testing. The guidelines translate that hard-won experience into a defined specification: a list of impurities, a maximum limit for each, and an analytical method to measure it. They are voluntary in the sense that ISBT is not a regulator, yet in practice they function as the commercial baseline; a producer who cannot demonstrate compliance will struggle to sell into the beverage market.
The structure: three reasons a limit exists
Every limit in the guidelines carries a rationale, and understanding the rationale explains why some numbers are far stricter than others. ISBT groups the reasons into three categories. Sensory limits protect the taste, appearance and odor of the finished drink; these dominate the list, because the consumer notices an off-flavor long before any safety threshold is reached. Process limits define parameters that matter to a controlled production process. Regulatory limits are set by governing agencies and are non-negotiable, because they concern safety rather than preference.
The headline limits
The guidelines specify CO₂ purity at a minimum of 99.9% by volume, with the balance made up of trace impurities held to tight ceilings. The table below summarizes the parameters that matter most in day-to-day verification.
| Parameter | ISBT limit | Rationale |
|---|---|---|
| CO₂ purity | 99.9% v/v min. | Process |
| Moisture | 20 ppm v/v max. | Process |
| Oxygen | 30 ppm v/v max. | Sensory |
| Acetaldehyde | 0.2 ppm v/v max. | Sensory |
| Aromatic hydrocarbons (as benzene) | 0.020 ppm v/v (20 ppb) max. | Regulatory |
| Total volatile hydrocarbons (as methane) | 50 ppm v/v max. | Sensory |
| Total sulfur (as S) | 0.1 ppm v/v max. | Sensory |
| Ammonia | 2.5 ppm v/v max. | Process |
| Nitric oxide / nitrogen dioxide | 2.5 ppm v/v max. | Sensory |
| Phosphine | 0.3 ppm v/v max. | Regulatory |
The two regulatory lines deserve attention. Aromatic hydrocarbons, measured as benzene, are capped at 20 parts per billion, and phosphine carries its own ceiling. These are the limits set on safety grounds, and they are the ones an auditor will scrutinize most closely.
Why benzene is the limit that defines the rest
Of all the impurities on the list, benzene is the one that concentrates the mind. It is a known carcinogen, it is regulated rather than merely undesirable, and at 20 ppb its limit is among the tightest in the guidelines. That combination makes benzene a useful proxy for the seriousness of a monitoring program: a system that resolves benzene confidently, with margin below the limit, is generally a system you can trust across the rest of the specification. A system that can only just reach the limit leaves no room for the natural variation in any measurement, and no room for confidence.
This is the practical heart of the guidelines. Meeting a limit on paper is not the same as proving it on a live stream, batch after batch. The value of a monitoring program lies in the margin between what the instrument can detect and the limit it is measuring against, a point explored in our companion article on detection limits and compliance limits.
How the specification is verified
The guidelines pair each limit with a defined analytical method, because a number means little without a repeatable way to measure it. Aromatic hydrocarbons, for instance, are determined by gas chromatography. It is worth noting that the ISBT limit is expressed "as benzene" and treats the aromatics as a single group; a system that speciates them resolving benzene, toluene, ethylbenzene and xylene (BTEX) individually rather than reporting one combined total. This tells you not just that aromatics are present but which ones, which often points to the source of a contamination. In practice, robust verification means measuring many impurities continuously rather than sampling occasionally, alarming on any excursion, and producing an auditable record for batch release. The strongest programs cross-check critical results with more than one analytical technique, so that a pass is corroborated rather than asserted particularly for the safety-driven limits where the cost of being wrong is highest.
What good looks like
A producer or bottler working to the ISBT guidelines should be able to answer three questions without hesitation: which impurities are being measured, what margin sits between the result and each limit, and how that result would stand up to an audit. A monitoring system that exceeds the ISBT guidelines measuring well beyond the minimum list, resolving the safety-critical aromatics far below 20 ppb, and issuing a secure certificate of analysis turns those three questions from a worry into a routine.
Learn more: see how AirBreather verifies benzene in beverage-grade CO₂ to 0.5 ppb, roughly 40× below the ISBT limit → [link to pillar page].
FAQ
- Are the ISBT CO₂ guidelines mandatory? ISBT is not a regulator, but the guidelines are the commercial baseline for beverage-grade CO₂; compliance is effectively required to sell into the market.
- What is the ISBT limit for benzene? Aromatic hydrocarbons, expressed as benzene, are limited to 0.020 ppm v/v (20 ppb), on regulatory grounds.
- How is beverage-grade CO₂ tested? Each impurity has a defined method; the most reliable programs measure continuously and corroborate safety-critical results with more than one technique.