Packaging and Carbonation: Bottle, Can, Keg

Beer is, among other things, a solution of carbon dioxide held in tense equilibrium with the liquid it flavors. The package — glass, aluminum, or stainless — is not a neutral container but a participant in that equilibrium, shaping how the gas behaves, how oxygen sneaks in, and how long the contents resemble what the brewer intended. Once a brewer accepts that the package is part of the recipe, a great many otherwise mysterious quality problems start to make sense.

A small physical chemistry digression

Carbonation in beer is governed by Henry's Law, which says, in plain terms, that the amount of a gas dissolved in a liquid is proportional to the partial pressure of that gas above the liquid. Raise the pressure of CO2 in the headspace, more dissolves; drop it, the gas escapes. Temperature matters in the opposite direction — colder liquid holds more gas at a given pressure, which is why warm beer foams violently when opened and cold beer keeps its head together.

Brewers measure carbonation in volumes of CO2, where one volume means one liter of CO2 (at standard temperature and pressure) dissolved in one liter of beer. The Brewers Association Draught Beer Quality Manual gives typical American ales a target around 2.4 to 2.8 volumes; British cask ales sit lower, around 1.0 to 1.5; Belgian styles and German weissbiers can run up past 4.0. These numbers are not stylistic affectations. They determine the pressure the package must hold, the temperature at which the beer pours correctly, and the geometry of the dispense system that follows.

Carbonation arrives by one of two routes. Natural conditioning uses residual or added fermentable sugar, plus live yeast, to generate CO2 inside the sealed container — the classical method for bottle-conditioned ales and traditional cask. Forced carbonation injects CO2 from a cylinder into cold beer under pressure, which is faster, more precise, and the dominant industrial approach. Each route produces CO2 that is, chemically, identical; the differences people notice in the glass come from yeast autolysis, residual sediment, and the trace gases dissolved alongside the carbon dioxide.

The bottle

Glass is the oldest of the three formats and, for a long time, the only one that worked. Its virtues are real: glass is essentially inert, contributes nothing to flavor, and tolerates the internal pressures of bottle conditioning without complaint. A standard 12 oz crown-finish bottle is rated well above the roughly 30 to 40 psi a highly carbonated beer might generate at warm storage temperatures, with a healthy safety margin engineered in.

The crown cap, patented in 1892, seals against the lip of the bottle through a compressed liner — historically cork, then PVC, now usually a tin-free PVC-free polymer in food-contact use. The seal is excellent but not perfect. Oxygen permeates through the liner at a slow but measurable rate, and the headspace trapped under the cap at filling contains its own oxygen budget unless the filler purges it with CO2 or a brief foam-on-foam fob. Peer-reviewed work indexed through PubMed Central on beer staling traces most flavor degradation in packaged beer to oxidation of unsaturated fatty acids and the appearance of trans-2-nonenal, the compound responsible for the cardboard note in old beer.

Glass has two well-known failings. The first is light. Ultraviolet and short-wavelength visible light cleave the isohumulones from hops, and the resulting radicals react with sulfur compounds to form 3-methyl-2-butene-1-thiol, a molecule the human nose detects at parts-per-trillion concentrations and recognizes, with some accuracy, as skunk. NCBI's review literature on hop bitter acids lays out the photochemistry in considerable detail. Brown glass blocks most of the offending wavelengths; green glass blocks some; clear glass blocks essentially none, which is why beers in clear bottles are either drunk quickly, brewed with chemically modified hop extracts that lack the vulnerable bond, or simply accepted as lightstruck by their drinkers.

The second failing is weight. Glass is heavy, which raises shipping costs and carbon footprint, and it breaks, which is unhelpful at swimming pools, beaches, and other places people enjoy drinking outdoors.

The can

The aluminum beverage can is, despite first impressions, a remarkable piece of engineering. A modern two-piece can is drawn from a disc of aluminum alloy, ironed thin enough that the wall is around 0.1 mm thick at the sidewall, and lined with a sprayed polymer — historically epoxy-based, more recently in many cases bisphenol-A-free alternatives — to keep the aluminum from contacting the beer. Without the liner, the chloride in beer would pit the metal and the metal would, in return, contribute a distinctly metallic note.

For carbonation, cans are excellent. The seam is hermetic in a way crown caps are not, light transmission is zero, and the headspace can be tightly controlled at the seamer. The Brewers Association's best practices materials note that total package oxygen, the metric brewers actually care about, is often lower in a well-run canning line than in a comparable bottling line, mostly because of the geometry of the fill and the speed at which the lid is seamed on.

The practical implications for a working brewer are several. A can is, structurally, a pressure vessel rated for normal beer carbonation but not unlimited carbonation; very highly carbonated beers, such as some Belgian styles approaching 4 volumes, push the design envelope and benefit from heavier-gauge ends. A can also requires a different filling philosophy — the beer must arrive at the filler already at its target carbonation, since there is no equivalent of bottle conditioning in a standard can. Some breweries have experimented with can-conditioning, dosing yeast and priming sugar before seaming, and the practice works, though the resulting sediment behaves a little differently when poured because the can geometry does not allow the drinker to see what is happening.

For the trained drinker, the relevant point is that a can poured into a glass is, by every measurable indicator, indistinguishable in flavor from the same beer in a brown bottle stored under identical conditions. The can-versus-bottle preference is a cultural and tactile question, not a chemical one.

The keg

A keg is, in effect, a very large can with valves. The standard half-barrel in the United States holds 15.5 US gallons, which by the federal definition of a barrel at 31 gallons (see 27 CFR Part 25 for the regulatory framework around beer measurement) is exactly half. A sixth-barrel — sometimes called a sixtel — holds about 5.16 gallons, and the so-called Cornelius or "corny" keg, originally built for soft drink syrup, holds 5 gallons even.

Stainless steel is the material of choice for nearly all commercial kegs because it is durable, sanitizable at high temperatures, and chemically silent. The valve at the top, a Sankey D coupler in most US applications, isolates the beer from atmosphere and connects to the dispense system through two paths: gas in, beer out.

The physics here are the same Henry's Law equilibrium as in the package, but now the brewer or operator gets to choose the headspace pressure. The Brewers Association Draught Beer Quality Manual devotes considerable space to what it calls applied pressure — the pressure of CO2 (or a CO2/N2 blend) maintained on the keg during dispense — because if the applied pressure does not match what the carbonation level and temperature equilibrium demands, the beer will either gain or lose CO2 over time. Too low, and the beer goes flat in the keg over a few days. Too high, and it foams uncontrollably at the tap and gradually over-carbonates. The classic chart relating volumes of CO2, beer temperature, and equilibrium pressure is the working brewer's most-consulted reference.

The dispense line itself is part of the system. A line that is too short or too wide does not drop enough pressure between keg and faucet, and the beer arrives at the glass in a foamy panic. Too long or too narrow, and pours are slow and the beer warms in the line. The Draught Beer Quality Manual specifies line resistance values, typically expressed in pounds per foot, and recommends balancing the system so that pressure delivered at the faucet is essentially atmospheric while the keg stays at its equilibrium pressure.

Kegs also bring a contamination consideration that bottles and cans do not. They are returned, opened, cleaned, sanitized, and refilled, sometimes hundreds of times. The Master Brewers Association of the Americas and Brewers Publications both maintain technical literature on keg cleaning, particularly the importance of caustic and acid cycles, internal spear inspection, and the periodic replacement of O-rings, which harbor the kind of beer-spoiling lactobacilli and pediococci that survive surprisingly hostile conditions.

Mixed-gas dispense and a small note on nitrogen

Stout dispense, and increasingly some other styles, uses a blend of CO2 and nitrogen — typically 75% N2 / 25% CO2 in the so-called beer gas mixture. Nitrogen is much less soluble in beer than CO2, which means the gas blend can be applied at higher pressures (useful for long draw lines) without over-carbonating the beer. The creamy, tight head on a properly poured nitro stout comes from the beer being forced through a restrictor plate at the faucet, which mechanically breaks dissolved gas out of solution as very small bubbles. The Brewers Association's draft quality literature treats this as a separate plumbing system, and most pubs that pour both standard and nitro draught beers run two parallel gas supplies.

The cask, briefly, for completeness

Across the Atlantic, the Campaign for Real Ale (CAMRA) maintains a definition of real ale that requires conditioning and dispense from the cask without added gas pressure — the carbonation is whatever the secondary fermentation produces, typically around 1.0 to 1.5 volumes, and dispense is by handpump or gravity. Cask is not so much an alternative format as a separate philosophy, in which the drinker accepts a shorter shelf life (a cask, once vented, is at its best for a few days) in exchange for a softer carbonation that lets certain malt and hop aromatics speak more clearly. Most working brewers in the United States will encounter cask only as an occasional festival exercise; in the United Kingdom and parts of Belgium it remains a daily reality.

What this means at the bench

For the brewer, the package decision is upstream of nearly every other quality decision. Targeting 2.6 volumes of CO2 in a hazy IPA destined for cans means a different filler setup, a different headspace strategy, and a different shelf-life expectation than the same beer destined for bottles or for a half-barrel keg pouring through 8 feet of 3/16-inch line. Picking up cardboard notes in a six-month-old bottle, or a flat last pint from a keg that has been on tap for three weeks, or a foamy first pour after a line cleaning — each of these is a story about gas, oxygen, temperature, and time, and the package is the stage on which the story plays out.

For the trained drinker, including candidates studying for the Certified Cicerone® exam or the BJCP exam, the useful posture is to ask, on tasting any beer, what it has been through to reach the glass. The answer almost always involves the package, and the package almost always involves Henry's Law.

Packaging and Carbonation: Bottle, Can, Keg

A small physical chemistry digression

The bottle

The can

The keg

Mixed-gas dispense and a small note on nitrogen

The cask, briefly, for completeness

What this means at the bench

Further reading

Read Next

Packaging and Carbonation: Bottle, Can, Keg

A small physical chemistry digression

The bottle

The can

The keg

Mixed-gas dispense and a small note on nitrogen

The cask, briefly, for completeness

What this means at the bench

Further reading

Read Next

Related Authorities