Mysterious phase-shifting liquid leakage
August 1, 2020 8:03 AM   Subscribe

I made ginger beer in a reused two-liter soda bottle, leaving less headspace perhaps than was wise. The next morning about 80% of the liquid had escaped, making a terrible mess, but the bottles were still whole and pressurized. What happened?

I used a pretty standard recipe: a cup of sugar, half teaspoon of cream of tartar, 1/3 cup of shredded ginger, 1/3 cup of lemon juice, raised to the boil, powders dissolved and then taken off the boil. Topped off with cold water after a period of cooling to 2 liters, a teaspoon of baker's yeast added when cool, open-fermented for 3 hours, and then strained into a bottle. I ended up (due to imprecision in measurement) with a bit more liquid than I ought to have, and topped the bottle up to within about 3/4 of an inch of the top --- probably a mistake, since that meant not much headspace. I carefully vented it midday yesterday, but this morning it was standing in a puddle of most of its contents, with strain marks (but no visible tears) at the neck. All this would have been perfectly explicable as a containment failure (notwithstanding the lack of obvious ruptures) expect that the bottle was still pressurized: it was rock hard, and when I turned the cap a little, it vented enthusiastically. What sort of failure mode leaks liquid but not gas?
posted by jackbishop to Food & Drink (8 answers total)
 
I have a hypothesis as to what happened. Have you ever used a pressure cooker before? When you put it on the stove and the pressure builds up, there is usually a small release at the top that allows for excess pressure to be released. Just because this pressure is being released doesn't mean there isn't still pressure inside the cooker. In fact if you foolishly tried to open the cooker during this process hot food would come flying out of the cooker at you. So my guess is that the liquid AND gas came out last night due to a build up of pressure. That doesn't mean that when it stopped leaking there still wasn't pressure in the bottle. In fact it was probably just under the amount of pressure that forced the bottle to leak. The pressure probably stopped building up when the liquid came out because all the yeast and sugar came out with it, so there was nothing to make new gas.

Does this fit your scenario?
posted by Toddles at 8:29 AM on August 1 [1 favorite]


perhaps it could have deformed under pressure to enthusiastically vent liquid & gas via the cap, and once pressure reduced, the bottle elastically relaxed back to a form that can hold some pressure and continued fermenting the remaining liquid building some pressure once again.
posted by TheAdamist at 8:35 AM on August 1 [6 favorites]


Deformation making the cap into something similar to a pressure relief valve, mentioned above, sounds likely. Another, perhaps less likely, possibility might be thermal changes. There's a good chance the bottle is PET and the cap is HDPE. HDPE typically expands more with heat. (The exact values span a very wide range, depending on specific details of the plastic and its history, with a difference in length roughly in the range of 2*10^-5/K to 2*10^-4/K.) That generally means the lid will become less tight as the bottle heats up. If you take the higher end of that range, going from tap water to body temperature would give you a 0.15mm difference across the diameter of the cap. Usually that difference gets taken up by stretching of the material when you screw the cap on tight. But, if there's a big change in temperature or the cap wasn't screwed on very tight to begin with, it could lead to a leak when warmed by fermentation that goes away once the bottle cools down again.

As someone who's had a high pressure 2L bottle explode dramatically in my hands - in my case while making very high pressure soda rather than brewing - I recommend safety glasses. And not re-using this particular bottle. (Also, I look forward to trying out your recipe.)
posted by eotvos at 9:31 AM on August 1 [1 favorite]


What sort of failure mode leaks liquid but not gas?

Tiny crack in the base, perhaps, that eases open only just wide enough to seep under very high internal pressure.

Any failure involving the cap shifting or stretching would pretty much have to have resulted in a sticky trail down the side of the bottle to the puddle underneath. If that's what happened I would have thought you'd have seen or smelt or felt it and wouldn't have been prompted to ask the question.
posted by flabdablet at 10:45 AM on August 1 [1 favorite]


I think it's possible and even likely that this failure mode is designed into the cap and bottle system.

You really wouldn't want the cap flying off the bottle like a bullet and hitting somebody in the face, or for the bottle to explode if you could avoid either of those outcomes.

I've noticed that the threads of the caps on my sparkling water bottles are not continuous but have multiple gaps. I wondered why they did it that way because it seemed more difficult to manufacture than a continuous thread. It didn't occur to me that the gaps could function as pressure relief vents until you asked this question.
posted by jamjam at 10:58 AM on August 1 [1 favorite]


I don't believe pressure relief is in fact a standard feature of disposable plastic soda bottle closures. If it were, there would be no need for this product and it would also be impossible to use those bottles to construct dry ice bombs, which I can assure you from personal experience do go off with a highly satisfactory earth-shattering kaboom.
posted by flabdablet at 12:36 PM on August 1


I think Mentos bombs were probably a stimulus for the development.

Dry ice bombs might be trickier because water ice could block a vent system
posted by jamjam at 12:42 PM on August 1


Home brewer here. Yeast eat sugars and produce CO2. This is how your beer (or ginger beer) gets carbonated, the CO2 produced by the yeast is trapped in the sealed bottle. As long as there is still some sugars present for the yeast to eat, they will keep eating sugar and producing CO2. Once the sugar supply is exhausted, the yeast will go dormant, but will reactivate when sugar is re-introduced.

This is how most home brewers carbonate their beer. The yeast is added to the wort and allowed to happily bubble away and ferment until it has "crashed out" and consumed all the available sugars, indicated by no more CO2 production. (At this stage, most brewers use an airlock, and CO2 production is easily seen from bubbles produced in the airlock. No more bubbles=no more CO2 production.) The beer may be secondary fermented and aged after this, but just prior to bottling, a carefully measured amount of sugar is re-introduced to the beer, and it is immediately bottled. This is just enough sugar to carbonate the beer, and then the yeast go dormant again, having exhausted their food supply.

Your ginger beer never crashed out, it still had tons of food available to the yeast when it was sealed in the bottle. The leakage from the cap is from excess pressure (and insufficient headspace) in the bottle forcing the liquid out. As there was still tons of sugar available, even after some of the contents had been lost, the yeast were still happily consuming sugars and producing more CO2, which probably forced more liquid out, and so on. The bottle, now missing some of its contents, was still pressurized because the yeast were producing CO2 all along, pressurizing the bottle, until they finally ran out of sugar to eat and took a nap.

Side note: If you are planning to consume your ginger beer immediately, it's probably fine, but the fact that your cap failed before your bottle burst indicates a not-great seal, and not only could this eventually result in a flat ginger beer, but once it goes flat or the pressure inside the bottle equalizes with the ambient air pressure, it could allow oxygen or other contaminants into the bottle, making for off-flavors, skunking, etc.
posted by xedrik at 9:36 AM on August 2 [1 favorite]


« Older BigBlueButton as a service, on demand?   |   Book recommendations for sensitive interiority Newer »

You are not logged in, either login or create an account to post comments