# Is It Possible To Calculate The Internal Pressure Which A Sealed Glass Vessel Can Support, By Creating A Relative Vacuum Outside The Vessel?

Questions are bold with context surrounding them.

I am not an engineer. Well, I'm a computer scientist but I don't do this. I do, however, brew kombucha and am looking to scale up. I have a new process that in theory will unlock better scaling, but I need to know how much pressure can build inside a sealed glass vessel with a 1 gallon capacity. It is hard to find glass vessels online with information on this. I do, however, have a vacuum chamber and pump. I wondered, if I seal a test vessel at atmospheric pressure, put it in the vacuum chamber and create a negative pressure around it, does that effectively simulate a positive internal pressure from inside, as if a fermented beverage was creating pressure?

I would aim for a vessel to be able to support at least 50 PSI but some leeway up to 70 would be great. So I would need to find out, what should be the pressure reading in my vacuum chamber to simulate a specific positive pressure within the sealed vessel? How do I calculate what the negative pressure needs to be to simulate 50-70 PSIG inside the vessel when the actual pressure without a vacuum outside is just atmospheric pressure?

Hopefully I'm using good terminology and this all makes sense. If something sounds wrong but in context you think you know what I mean, feel free to make corrections which will help me learn.

• Aside from the fact that you only get about -14.5 psi at the extreme (depending on altitude and weather conditions), vacuum pumps have to be protected from sucking stuff into them (I.e. debris from your specimen, should it break during the test), so pressure systems tend to be cheaper Commented Apr 13 at 5:37
• So using this method I won't be able to test for any more than almost 1 atm before I need a mechanism for pumping pressure into the jar anyway. Though the vacuum method would still be useful with that limit as I can put a balloon in the vessel and also check for leaks (if I hold a vacuum and the balloon starts to expand then there's a leak, which is another thing I've been wanting to test simply and cheaply). For the record I have screens I could put in the chamber to protect anything other than potentially ultrafine particulates from getting into the pump. Commented Apr 15 at 7:51

The lowest pressure you can get is zero absolute. Air pressure is about 15psi absolute. The greatest difference you can get across the glass with a vacuum is then only 15psi.
If you want to test your glass container, get an air pump with a pressure gauge, find some way to seal your container with a nipple attached and pump it up. Do this with the container inside a larger can, preferably with a lid, to protect yourself if it does explode.
Do you know what pressure you're likely to develop during fermentation? There should be some info out there. Have other people had bottles explode on them? How likely is this to occur?
I know that when I was in college my roommate brewed a lot of beer in regular bottles (in his closet) only to have them all explode one night, one after the other. A big mess.

If you pressurized your vessel to 30psi, and then pulled a vacuum around it, you'd be testing it to 45 psi.

Edit:
Testing with water rather than air as suggested by Eric S:
If one gallon of water is compressed to 60 psi the energy in that water is hardly measurable. If you have one gallon of air at 60psi there is about 20 Joules of energy. That's equivalent to the energy of one pound weight falling 12 feet. Not something you'd want to have hit your foot. EricS, talking about scuba tanks, then there is 50 times more energy, and that's a real issue. If you do as you suggest and fill the bottle with water and then pressurize it there will be only the energy of the air in the hose which is considerably less.

You could also check the pressure of your local water supply, that generally varies between 50 and 80 psi.

IMPORTANT Forgot to add. The stress on a pressure vessel increases in proportion to the radius. This means that is you double the radius you'll need glass of twice the thickness to support the same pressure. You may be safer to stick with bottles rather than gallon containers.

• It is much, much safer to test a pressure vessel with a pressurized liquid like water than a gas like air. Commented Apr 13 at 1:08
• Now this is handy! (The knowledge of air pressure being 15psi and its implications to my suggested tested method) I was predicting that to be a potential problem. Thank you for helping clarify. I have been making kombucha for some time now at smaller volumes where the pressure builds in smaller pint bottles designed for kombucha. I have had very carbonated batches but have never had an explosion. However, I have heard of them from kombucha and beer brewers alike. The expected pressure is similar to beer, maybe a little less than champagne. More at room temperature but lessens when cold. Commented Apr 13 at 1:12
• @EricS How is that? Is the air more explosive because of the rapid decompression whereas liquid can't be compressed? Is a 60psi rapid decompression that dangerous (with protection)? I could certainly follow Rich's suggested testing method in a larger steel cylinder for protection. If I were to use liquid could I simply fill it most of the way with liquid and then pump in air until 60psi all the same? Maybe less compressed air means less of a pop? Commented Apr 13 at 1:17
• Absolute vacuum is 0 psia. Atmospheric pressure is 14.7 psia. The largest pressure differential you can get by pulling a vacuum is 14.7 psi. The safe way to test a pressure vessel is with a pressurized liquid like water. That is what people do with scuba tanks which can support 3000 psi. Commented Apr 13 at 1:20
• Pressure for hydrotest can also be very low tech, like a very long small diameter tube, you just need to raise the end with free surface. This way you get about one atmosphere for every 10 m. Commented Apr 13 at 20:32

If you want to simulate 60 PSI internal pressure, use 60 PSI external vacuum. It's the same thing.

Glass exhibits high variability in safe working pressure. Always assume pressurized glass bottles will explode: use safety glasses, protective gloves, and protective footwear (preferably with insole protection).

An exploding bottle can still cut into your arms and legs, but your eyes are generally considered much more critical. The gloves are for handling: if the bottle explodes in your hand, it's your hand that will get cut. And the safety shoes are because when the bottle explodes in your hand, it will drop on your feet, and after the bottle explodes, there will be broken glass on the ground.

Champagne bottles are particularly designed to handle pressure, but they still sometimes explode in handling. They are handled with care, kept in temperature regulated conditions, and by the time they get to the table, the weak ones have mostly already failed.

Workers who work with pressurized bear and wine bottles experience failing bottles all the time, and take precautions. If you're going to work with pressurized glass vessels, you should do the same.

• Since testing this brewing method I've been extra careful, ramping up fermentation times and wearing protective gear. Thanks for the advice. Commented Apr 13 at 1:06
• What is a 60 psi external vacuum? The best you can do is -14.7 psi from atmospheric pressure. Commented Apr 13 at 1:11
• @EricS That makes a lot of sense, using water. The amount of energy stored in the water is minuscule compared to that of air as water is basically incompressible.
– Rich
Commented Apr 13 at 1:29
• If one gallon of water is compressed to 60 psi the energy in that water is hardly measurable. If you have one gallon of air at 60psi there is about 20 Joules of energy. That's equivalent to the energy of one pound weight falling 12 feet. Not something you'd want to have hit your foot. EricS, talking about scuba tanks, then there is 50 times more energy, and that's a real issue. If you do as you suggest and fill the bottle with water and then pressurize it there will be only the energy of the air in the hose which is considerably less.
– Rich
Commented Apr 15 at 13:16
• @Rich, I hadn't really thought about that, but yes, that explains why we never worried about cutting through our clothing: the explosions we saw were more like bursting bubbles: a spray of broken glass, but not much punch. The gloves, safety glasses and insole-protection were real though. Commented Apr 17 at 3:23