Canned Pale Ale Staling While Kegs Stay Fresh
A 20-barrel canning run of a 5.0% hazy pale ale finished filling 2,400 twelve-ounce cans. The QA sheet shows dissolved oxygen at 8 ppb in the bright tank, 45 ppb in the filler bowl, and 110–180 ppb in canned product across six samples; the target is under 50 ppb. The line ran at 28 cans per minute against a rated 40, with two short stops. A seam teardown passed. Eight weeks later, accounts report cardboard and sherry notes in cans while the same beer in kegs still tastes fresh. The brewery switched to a new bulk CO2 supplier two months ago, and the lids are from a new lot.
1. Interpret the dissolved oxygen numbers. Where is pickup occurring, and how much does 110–180 ppb matter?
Read the worked answer
Reading the numbers.
- Bright tank at 8 ppb is excellent, so the beer left the tank in good condition.
- Filler bowl at 45 ppb shows pickup between the tank and the filler - transfer line, pump, or inadequate purging of the bowl and its headspace. Something in that path is adding about 37 ppb.
- Package at 110–180 ppb is two to four times the 50 ppb target, showing a second and larger pickup at the filler, the lid, or the can itself.
How much it matters. 150 ppb in a 355 mL can is about 53 micrograms of oxygen - small in absolute terms, but stale-flavor aldehydes such as trans-2-nonenal are perceptible at extremely low levels, and the reaction rate roughly doubles for every 10 °C of storage temperature. Eight weeks is plenty of time for cardboard and sherry notes to develop, especially if the cans were stored warm. That the kegs stayed fresh tells you the package, not the beer, is the problem - kegs were presumably filled on a different line or with better purging, and a keg has far less headspace oxygen per litre.
First, verify the measurement. One hundred-plus ppb is high enough that a sampling error, a poorly purged piercer, or a drifting oxygen sensor can produce a false reading. Confirm calibration and sampling technique before spending money on the line.
2. Rank the likely sources and describe how you would confirm each one.
Read the worked answer
Ranked suspects, highest first.
- CO2 quality. The supplier and delivery method changed two months ago - the same window as this run. Bulk CO2 with elevated oxygen contaminates both carbonation and every purge on the line. Action: request the certificate of analysis, confirm beverage-grade spec, and test the gas at the tank outlet with an oxygen analyzer or a gas purity kit. Do this before anything else; nothing on the filler matters if the gas is dirty.
- Filler purge and counter-pressure. Insufficient pre-purge of the can, weak CO2 counter-pressure, or short purge times leave air in the can before filling. Confirm with a pickup study: DO at the bowl, after can purge, and after fill, on the same can position across the run.
- Line speed and stops. Running 28 cpm against a rated 40 means each can sits longer in the filler, and the two stops left beer standing in the bowl and filler heads. Confirm by running a controlled speed test at 40 cpm and comparing package DO, and by recording DO before and after a deliberate stop.
- Lid lot and seam. The seam passed teardown, but a new lid lot can bring different compound, coating, or handling. Confirm with a lid-swap trial against retained lids from the previous lot, and re-check seam dimensions on the new lot.
- Foam and fill height. Insufficient foam at seaming leaves headspace air; over-foaming causes fobbing and low fills. Confirm by measuring fill height and headspace on the same cans you test for DO.
- Transfer path to the bowl. Confirm the 45 ppb by testing at the bowl inlet, and check pump seals and line purging.
Sequence: one variable at a time, testing package DO after each change, and keep a control sample of the current cans.
3. It is week eight and accounts are complaining. What do you do with the remaining pallets and the account relationships?
Read the worked answer
Quality decision. The defect is already evident in trade, so the product is out of specification on flavor regardless of what the DO meter says. Hold the remaining pallets - do not ship them - and pull what is still in accounts. Do not discount it into the market; a cardboard-tasting can is worse advertising than an empty tap handle.
Investigate before disposing. Open a statistically sensible sample: cans from several lots and from cold versus warm storage, plus a retained keg sample. Run a triangle test against a fresh control and, if possible, DO/TPO on the cans. This tells you whether every pallet is affected or only the warm-stored portion, and it gives you evidence for the CO2 supplier conversation and for any insurance or supplier claim.
Communicate. Tell accounts plainly that there is a package-oxygen problem in cans, that kegs are unaffected, and give a firm date for replacement cans off a corrected line. Offer kegs or a credit in the interim. Handled quickly, this costs a few cases of beer; handled slowly or evasively, it costs the account.
Prevent recurrence. Keep the DO pickup study as a permanent record on every canning run, add a package DO release limit (under 50 ppb, with any exceedance triggering a hold), require a CO2 certificate of analysis with each delivery, shorten the code date for cans, and ship and store cold. Also review whether cans should be released at all until the line is requalified.