Ask Mr. Wizard

CO2 Flushing

TroubleShooting

Bill Serowski - Hamburg, New York asks,
Q

I always hear that CO2 is heavier than air, but I am no scientist. Just how much heavier than air is it? I would like to know if the method of injecting a little CO2 into a carboy before transferring or blasting a little into a bottle before bottling has any impact on keeping the oxygen out of your finished beer? I have a vision in my head of this blanket of CO2 sitting in the bottom of a carboy or a bottle, then filling them from the bottom up where the beer nicely stays under the blanket of CO2 and eventually all the oxygen gets pushed out as the level rises in the carboy or bottle. Is this really what is happening?

A

Oh boy, this topic is one that I have some pretty strong thoughts about! I will start by throwing out a few numbers to answer your question about gas weight and density. Carbon dioxide weighs 44 grams/mole, oxygen weighs 32 grams/mole, and nitrogen weights 28 grams/mole. Since air is about 79% oxygen and 21% nitrogen, the weighted average comes out to be 31 grams/mole. According to the Ideal Gas Law, an “ideal gas” occupies 22.4 liter per mole at atmospheric pressure and 0 °C (32 °F). I will leave the discussion about how this assumption is problematic for engineers to physical chemists. The take home message is that there is certainly a difference in density between carbon dioxide and air, with air having only 70% of the density of carbon dioxide.

To put this density difference into perspective, ethanol has a density of 0.79 kg/liter, compared to 1 kg/liter for water. If you carefully pour ethanol on top of dyed water (using a handkerchief helps this experiment) you can easily float ethanol on top of water (the dye in the water helps visualize this effect). The same is true if you carefully “pour” carbon dioxide on the bottom of a container of air. The problem is convection.

Take your glass of dyed water with the ethanol cap and heat the bottom of the glass with a hot plate or heat source of your choice. The water will begin to move around because of convective currents and these currents cause some mixing to occur at the interface between the water and the ethanol. If the layer of ethanol is really thick you will end up with water on the bottom, a mixture of ethanol and water in the middle and ethanol on the top. If the ethanol layer is thin, the glass will completely mix due to convection. If you did the same experiment with oil and water, the oil would always float on top of water because the two liquids are immiscible.

Gases do the same thing when there are differences in temperature in the system, for example the carboy and the gas being added to it, and gases don’t behave like oil and water. This means that gases mix and tend not to layer. Add Brownian Motion to convection and what results is a system that will eventually equilibrate. Carbon dioxide can effectively be used as a blanketing gas if it is carefully added to the bottom of containers (carefully means low velocity to minimize mixing) and quickly used to guard against oxygen ingress into beer. In order to do this it is important to create the blanket and fill beer beneath it before the blanket is infiltrated by air.

So how can this be applied to homebrewing? For starters, if you do want to blanket a carboy or bottle with carbon dioxide you need to add the carbon dioxide to the bottom of the container. If you add it to the top of the container it will sink and mix with the air and never form a solid layer at the bottom. And if you use this method you need to diffuse the flow. A gas diffuser stone works great for this method. A cheap aquarium stone slipped onto the end of a gas stone will do the trick.

But how do you know that this works? Unlike the water/ethanol demonstration you cannot conveniently dye gas and there is always a level of uncertainty without the use of an expensive gas meter. My go-to method is to fill a container to the brim with water and then push the water out using carbon dioxide. While not practical for bottle filling, this method is both practical and very effective for purging kegs and beer serving tanks. We use this method to prepare our 500-gallon (1,900-L) serving tanks for filling at Springfield Brewing Company. The water we use is recovered, so the only thing our method requires is time to fill our tank with water and then to displace the water with carbon dioxide. Other breweries layer carbon dioxide into the bottom of their tanks.

I know our method seems pretty extreme, but I am not convinced that we use any more gas than we would if we attempted to flood the tank with carbon dioxide while displacing air in the process (a relatively common method). My day job with the Paul Mueller Company has exposed me to some really interesting processes over the last 18 years. We have built numerous million-gallon (3.8 million L) aseptic orange juice storage tanks over the years and that industry is very concerned about oxygen because oxidized orange juice takes on a brownish color and loses Vitamin C in the pro-cess. These huge orange juice tanks are flooded with 1 million gallons (3.8 million L) of iodophor, then pushed out with sterile nitrogen and completely drained prior to filling. This method not only displaces all of the air from the tank, it also ensures complete contact of the sanitizer with the tank wall.

Response by Ashton Lewis.