CO2 sensor

[TheBiGZ]

So after wasting about 50 pounds of CO2 over the past few years to leaking kegs, I started looking for a carbon dioxide sensor. Most places said, “don’t you mean carbon monoxide?” Ugh...
So having some experience with micro controllers, I decided to build CO2 meter. I ordered an infrared CO2 sensor from eBay and wired it up to an Arduino. Then send the reading via USB to a Raspberry Pi outside the fridge. It works great but there is an anomaly that I can’t quite figure out.
Normal atmospheric CO2 level is about 400 ppm so that’s what I’m looking for. Much higher than that means I have a leak.
So here’s the issue. Every morning, it’s about 395-405 ppm...exactly where I want it. Then, at various times throughout the day, the readings spike to 600 or 700...sometimes 800. For the first few days, I thought I had a leak, but the reading is always around 400 in the morning and I haven’t lost CO2. The fridge should be relatively sealed, right?
Any scientists out there able to tell me what’s happening? View attachment IMG_1158.jpgView attachment IMG_1157.jpg

 

[schematix]

Sub. Interested in using something like this for my basement where I ferment and brew.

 

[mongoose33]

What kind of temperature variation do you have? Is it possible that as things warm up (or cool down), a leak appears which goes away when the opposite temperature extreme is hit?

You say you haven't lost CO2 but how do you know that? My regulators don't tell me how much CO2 is left, just that I'm still at max pressure in the CO2 tank. You might be losing CO2 and until the liquid C02 in the tank disappears, the gauge will never catch that.

 

[MaxStout]

The best way I know of to determine if a cylinder is losing CO2.

 

[day_trippr]

It would indeed be interesting to plot the interval ppm reading against the temperature over the course of 24 hours...

Cheers!

 

[TheBiGZ]

What kind of temperature variation do you have? Is it possible that as things warm up (or cool down), a leak appears which goes away when the opposite temperature extreme is hit?

You say you haven't lost CO2 but how do you know that? My regulators don't tell me how much CO2 is left, just that I'm still at max pressure in the CO2 tank. You might be losing CO2 and until the liquid C02 in the tank disappears, the gauge will never catch that.

I’m not sure about the temp variation. The door stays closed so I’d assume it’s pretty steady.
I say I’m not losing CO2 because it always goes back to 400ppm. If I was losing it and I didn’t open the door, it wouldn’t go back down to atmospheric range.
I’m guessing when the compressor kicks on, air gets circulated so the air around the sensor changes. Then when it turns off, CO2 would settle to the bottom where the sensor is and would be high.
Does a fridge evacuate air too? Does fresh air ever get pulled in?

 

[TheBiGZ]

Until now...

 

[doug293cz]

...

I’m guessing when the compressor kicks on, air gets circulated so the air around the sensor changes. Then when it turns off, CO2 would settle to the bottom where the sensor is and would be high.

...

That's not it. CO2 does NOT settle to the bottom. Diffusion insures gas mixtures homogenize over time. Check out the video. Br2 is about 3.6 times heavier than CO2 and it homogenizes with air in about 30 minutes in the video.

[ame]https://www.youtube.com/watch?v=_oLPBnhOCjM[/ame]

Brew on

 

[mongoose33]

That's not it. CO2 does NOT settle to the bottom. Diffusion insures gas mixtures homogenize over time. Check out the video. Br2 is about 3.6 times heavier than CO2 and it homogenizes with air in about 30 minutes in the video.

https://www.youtube.com/watch?v=_oLPBnhOCjM

Brew on

I've often wondered if the reason people believe this is from having seen dry ice sublimate into very cold gas which then seeks a lower level. If it were not colder than the surrounding air, it wouldn't sink like that.

 

[TheBiGZ]

I've often wondered if the reason people believe this is from having seen dry ice sublimate into very cold gas which then seeks a lower level. If it were not colder than the surrounding air, it wouldn't sink like that.

Not sure I buy that completely. I have often performed the parlor trick where you fill a beer glass with CO2 and then pour it over a candle to extinguish the flame. Maybe they will homogenize over time, but wouldn’t the ppm increase and then stabilize regardless. I would never get back down to atmospheric ~400ppm.

I think it’s common knowledge that CO2 is heavier than air/oxygen...that’s why we purge our kegs. Are you saying that effect is only temporary?

 

[doug293cz]

Not sure I buy that completely. I have often performed the parlor trick where you fill a beer glass with CO2 and then pour it over a candle to extinguish the flame. Maybe they will homogenize over time, but wouldn’t the ppm increase and then stabilize regardless. I would never get back down to atmospheric ~400ppm.

I think it’s common knowledge that CO2 is heavier than air/oxygen...that’s why we purge our kegs. Are you saying that effect is only temporary?

You only have to drop the O2 content from the normal 21% to about 16% to put out the candle. So, if you pour relatively quickly, then the O2 content will still be low enough, even if some mixing with air occurs. Fill the glass with CO2, wait 15 minutes, and see if the candle still goes out.

If you fill a purged keg thru an open lid, you lose most of the benefit of the purge.

I suspect your sensor is giving you bad readings sometimes, but have no idea why it would.

Brew on

 

[mongoose33]

Not sure I buy that completely. I have often performed the parlor trick where you fill a beer glass with CO2 and then pour it over a candle to extinguish the flame. Maybe they will homogenize over time, but wouldn’t the ppm increase and then stabilize regardless. I would never get back down to atmospheric ~400ppm.

I think it’s common knowledge that CO2 is heavier than air/oxygen...that’s why we purge our kegs. Are you saying that effect is only temporary?

Well, I'm pretty sure we purge our kegs to get the air out of them, air that includes oxygen that contributes to oxidation.

I'm not saying anything about any effect being temporary or otherwise. My comment was purely about why people might think there's a "blanket" of CO2 on the bottom of a vessel into which CO2 has been released.

I'm simply wondering if people are remembering something like this (skip to 38 seconds):

[ame]https://www.youtube.com/watch?v=bNOol7Nh07Y[/ame]

and assuming that it would also be applicable to CO2 that isn't super cold. Actually, what we see isn't CO2 but water vapor condensed into a "fog" that, being cold, sinks through the warmer air surrounding it.

 

[MaryB]

One of your kegs has a weak PRV and is burping once a day as pressure builds? Maybe fermentation not quite finished?