Hack Your Sodastream With A Giant CO2 Canister

Sodastream machines are popular amongst people who like to make their own seltzer water at home. However, replenishing the tiny gas canisters is expensive and wasteful. [Becky] decided to upgrade her machine to avoid this problem, and added some smarts while she was at it. 

The simple part of the hack is using an adapter to connect the Sodastream apparatus to a 50 lb CO2 tank from the welding store. This is easy enough, and just uses a off the shelf adapter. Using welding-grade gas in your drinking water is probably a really bad idea, but [Becky] was willing to take the nisk.

However, safety was given due attention in that a CO2 monitor was installed to make [Becky] aware of any dangerous leaks. The tank is also placed on a custom scale built with load cells and an ESP8266, which allows monitoring of how much gas is left. [Becky] notes that at her rate of drinking one bottle a day, the tank should last her a full 7 years or so.

The project brings costs down to 18 cents per liter of seltzer, versus 38 cents for the Sodastream gas supply. It’s likely that the Sodastream prices could still be beat even if a food-safe CO2 source was used. Plus, there’s no need to regularly buy new bottles!

Overall, it’s a great project and one that recalls us of continuous-ink printer hacks. Video after the break.

83 thoughts on “Hack Your Sodastream With A Giant CO2 Canister

  1. Please, please, please don’t store a CO2 cylinder in your house, no matter how much you trust your alarm.

    The toxicity profile is very different to carbon monoxide, but an undetected leak overnight might just kill you. I’m an anaesthesiologist.

    1. According to the safety data sheet the stock cylinder is 57.3 bar. They attached a 50 lb gas cylinder pressurized at 124 bar without regulator straight to the soda device. I’m an staring in fear at my screen.

      1. Yes, but the original device used screw-in cylinders that would have almost certainly been filled to the point of having a 2-phase system (liquid + vapor), so the pressure it was designed for would have been the same. The fact that there’s more liquid in the larger tank wouldn’t affect the pressure. In a 2-phase gas/vapor system, pressure is determined solely by temperature. Google “Gibbs Phase Rule” for more info.

        https://en.wikipedia.org/wiki/Phase_rule#Pure_substances_(one_component)

        / chemical engineer

        1. so how come my Sodastream carbonating tube runs constantly when connected to a big tank? I even tried a brand new Sodastream and it did the same thing. It must need a regulator.

    2. A CO2 leak would be quite detectable since you’ll get a burning sensation when breathing, aren’t you thinking about nitrogen or carbon monoxide since those are the silent killers?

      Many people with aquariums also have these big cylinders stashed in the cabinet of their aquarium and I’ve never heard of anyone having issues with it.

      1. which part of “I’m an anaesthesoiologist” did you not understand? If you wake up (if you do) in a CO2 flooded room you will defnitely not make it out of that room. You’re in absolute choking death panic and you have no idea what is going on. Good luck with that.
        The recommended minimum room size with furniture to store a 10kg CO2 cylinder is > 75m² at 2.5m height.

        1. Ok chicken little, every homebrew enthusiast in the country will certainly follow your sky falling advice about CO2 cylinders. Which class for your degree covered the safety mechanisms and certification every cylinder goes through before being filled? None of them? That explains your fear mongering.

          1. JP: see references below

            (The several hundred hours of training I go through each year, on top of my degree, licenses, and cetrifications, in safety and procedure is certainly more than you have had for your home brew hobby. Not to mention the first post here is someone whose knowledge of the particulars of breathing assorted non-standard-air gasses trumps mine, and yours, by far. Really.)

        2. Are you really? .. frankly as a professional brewer I have been knocked on my but with CO2. Your out, like asleep, been picked up off the floor and watched people pass out in walk in coolers where all the lines are most likely to leak. Our alarms are set at 5000 ppm, which is well below most pot grow farms of 8000 ppm. I have had my 750 lb tank leak and it did not even set off the alarm. CO2 stays low and you would need to be low as it will flow like water.

          Now I would suggest most home users at a 5lb tank due to cost and easy to exchange.

          But what your describing I have never seen in all the years working in breweries.
          BTW, .. cleaning our tanks can set off the alarms for 5000 ppm

          https://images.app.goo.gl/vMcZkdZFbjAW54FV9

      2. CO2 does not cause a burning sensation. The effects are concentration dependent. Increased respiratory rate, confusion, sedation, then a counterintuitive fall in respiratory effort, respiratory arrest, death, Are you thinking about something else? Look up CO2 narcosis. I’ve seen it in the flesh many times (usually drug-induced).

          1. Any cold dry gas causes coughing. When any gas expands rapidly, it cools. If I were to give you 100% cold, dry oxygen sufficiently quickly, you will get some respiratory irritation. We humidify and warm all gases given to patients in high concentrations for this reason. There are also contaminants (hydrocarbons) in the cylinder, which may cause irritation,

            CO2 is not a direct irritant at room temperature and normal humidity.

          2. There is some suggestion that it can be a direct nasal irritant (causing pain) in awake subjects who get exposed to brief high concentrations. Following up from my other comment though, those subjects are being exposed to CO2 straight from a cylinder, which is cold and dry.

          3. What I experienced was perfectly room temperature.

            The point was to see if your voice got lower instead of higher if you substituted CO2 for helium in a party balloon and inhaled it. Nobody could inhale the CO2 because it would make people cough and gag instantly.

          4. I believe this is because when you put very high concentrations of CO2 around your mucous membranes, enough of it dissolves into the water there to drop the pH pretty aggressively, which causes irritation. That being said, I trust the anaesthesoiologist about what the effect of a leaking tank in your room would be.

    3. CO2 isn’t as dangerous as carbon monoxide — not even close. Humans can’t detect carbon monoxide, it’ll just silently poison you, but we’re *very* good at detecting excess CO2. Just hold your breath. The uncomfortable “I HAVE TO BREATH RIGHT NOW” feeling you get after a minute or so is from the CO2 building up in your blood.

        1. If you manage to breathe in. I have experienced breathing pure CO2 and it doesn’t instantly drop you unconscious – there’s a period where you start to “fuzz out” and you could still get out of the room if you leave immediately. It’s a horrible and painful experience though, and most people would panic and fail to get out.

      1. Welding gas is 100% the same as any other CO2. CO2 bottles have a CO2 specific valve so there is no cross contamination. Fill plants do not differentiate between welding and food.

        I worked as a welder repair tech in a building that was shared with a bottle fill plant.

          1. Yes–it is. There is only one giant dewar that fills all of the bottles. The little 5 pound bottles that go on your kegerator are filled right next to the 50 pound ones going to a weld shop.

            Bottles are pumped down with a vacuum pump and then refilled.

        1. The CO2 is from the same CO2 plant as food grad CO2. They don’t make different grades at source, it’s all the same. The difference between food grade and non at delivery is simply the canister used to hold the gas. That 50lb welding canister may have whatever left inside, so the gas is still CO2 but you may get a bit of oil or some other crap in it, or you may not. Nobody checks.

        2. I did some research on this and it can happen that they use the same pumps and most of the time its the same stuff, but it doesn’t have to be that way since the pump can use other grease and oil that you wouldn’t have in food grade stuff.

          1. I kind of doubt it. Medical bottles are guaranteed to have been only used for oxygen. Also smaller ones have a special bail type quick connection. But oxygen is oxygen and the people that got up in arms about getting industrial oxygen do not have a clue what they are talking about. Oxygen is “made” from distilling air.

          2. Because one is certified (ie, they’re criminally liable for it) and the other is not. Also oxygen is a bad example since it tends to rapidly oxidize any contaminants

          3. I think in Germany you aren’t even allowed to sell gas in the wrong colored bottle.
            There are some regulation regarding the least amount of pressure a bottle can hold, when I was a paramedic here in Germany, we weren’t allowed empty the tanks completely since that way they have positive pressure. If we would happen to empty one completely the tank had to get cleaned and disinfected, which was pretty expensive.

  2. Where I leave, CO2 fire extinguishers are also a good source of cheap CO2. Plus if you buy many units, the seller lower the price: one close to my 3D printer, one other in my car, and one last in my beer tap ;)

    1. Spent a few years filling and testing extinguishers, be sure to let them know it needs to be food grade, we had cold tanks from a supplier we could use, but we also tried our best to recycle most of the CO2 from tanks we emptied, every once in a while the gas leaking off while connecting would smell of death and would already be mixed into our re-use banks.

    2. Yes, but sometimes that gas leaves a metal-like taste in your soda.
      Usually it’s solved by cleaning the tank and going to a different charging station.
      At least in my country the company Drago has been selling both extingishers and soda gas for the last 50 years with the same gas.

  3. This definitely needs a short, hard-piped connection to an automatic shutoff valve under the control of some proper safety systems if you don’t want to risk gassing your family. That much CO2 in a domestic situation is a very bad idea.

    1. This happens a lot in the aquarium hobby. I have 10 kg tanks on mine because the bigger ones don’t fit in my cabinet but I know people with one or more B50 stashed in a closet in their living room or bedroom. Never heard of any incidents and if it is seriously leaking you will hear it and it burns while breathing so you won’t be surprised in your sleep.

  4. Nice to not throw things away, but soda water is under 18 cents a litre here from the shop. Not that we buy it often. The daughter went through a phase of liking bubbly water wnen she was 4.

    As to welding gas vs food grade – guessing a bit, but could be lightly contaminated with process gasses from which it was ectracted or oils from a non food grade compressor.

      1. No, 18 cents is the post-hack cost. 38 cents is the standard cost. From the article:

        “The project brings costs down to 18 cents per liter of seltzer, versus 38 cents for the Sodastream gas supply.”

        1. 38 cents is not “the standard cost”. The common cost to replace the Sodastream canister is about €10-15 which translates to 16-25 cents a liter. Off-brand replacement canisters come even cheaper.

    1. The inlet on the sodastream expects full bottle pressure, up to 700 Bar. Inline carbon filters that can withstand that are non-trivial.

      Generally the nut and tail that connects to the bottle has a built-in sintered particle filter.

  5. A 5lb CO2 bottle used in home brewing/kegging equipment is probably a better option:
    1) You can own them outright and not pay any ongoing bottle rental costs
    2) The refill costs are pretty good, not much worse than the 50lb bottles and offset by the rental costs. Both are still far better than the sodastream costs.
    3) Being smaller it’s less likely to harm you if it leaks. Simply not enough volume stored to fill a house.

  6. Sounds like she’s getting ripped off on her CO2. Here in Sweden you can get Sodastream bottles refilled for 20kr (2.32 $US). That’s supposedly good for 60 litres, making for less than 4c / litre.

    Bonus is you have no responsibility for the cylinder maintenance, and if something goes wrong you’ve lost a couple of bucks.

    1. It was £7 a bottle in the UK, before the CO2 shortage caused by the rise in gas prices.
      The wholesale gas price rose by a factor of 4m, and the company that makes a lot of the UK CO2 as a byproduct of making ammonia fertiliser, sold its contracted gas back to the market at a bigger profit than they could make from operating their plant. Then shut down the plant.

      So the UK was running out of CO2 last month which is also used for packaging foodstuffs to make them last longer… and ironically by breweries (!) to make the beer fizzy.

      1. Breweries use CO2 because the beer is microfiltered to remove yeast, so it cannot carbonate itself in the bottle/keg (and explode as a result). Larger breweries capture and re-use the CO2 from the fermentation, while smaller ones don’t because the equipment costs more money than simply buying it bottled.

  7. Awesome! We go through a little bottle in under a week, and I’ve often thought of doing similar, I’m not sure my wife wants a hole in the kitchen counter top though, or a massive bottle sitting up there…. :)

    1. Some people refill their Soda Stream bottles by unscrewing the valve then dumping dry ice pellets into it, then screwing the valve back on. Once the dry ice sublimates its ready to fizz up some water.

  8. Sodastream bottles are a piece of cake to refill using dry ice. Available from your friendly local industrial gas supplier in pellet form. There seems to be no specific warnings, the MSDS states “CO2 gas is not flammable and is without odour or taste. It is not toxic, approved as food additive.” ( https://static.prd.echannel.linde.com/wcsstore/AU_BOC_Industrial_Store/pdf/product/en_AU/Safety%20Advice%20-%20Handling%20and%20usage%20of%20dry%20ice%20BOC.pdf )
    I’ve got half a dozen emptys to fill one day. Thought they might make a useful energy fur summit sumfing or ova
    Link to method… https://www.youtube.com/watch?v=2qSwLIXKTHs

  9. You could save more money by not drinking soda, and save dental bills too!
    But, seriously, neat. Drives me up the wall when devices need tiny refills which are an order of magnitude more expensive than bulk.

  10. Umm…home brewers have been carbonating their home brewed beer and things like rootbeer or other flavored soda for decades with CO2 from the local welding supply shop. You don’t even need a Sodastream.

  11. I know it wouldn’t make such an interesting video, but the easier solution would just be to buy the adapter to refill the official gas bottle from the big CO2 tank. That way you don’t have to keep the tank in the house.

  12. SodaStream and other carbonation systems like it are not a very good choice.

    If you’re going to get a CGA320 bottle anyway, it’s better to get a regulator and hoses, ball lock attachment and a carbonation cap. You can get dirt cheap plastic carbonation caps from KegLand now that work great, or spring for stainless ones.
    You can not only carbonate any beverage you want, including things not recommended in a SodaStream, you can dial in the pressure you want to carbonate at. It also gives you almost everything you need if you want to move up to kegs after. It’s also fairly easy with a keg to make a continuous draw soda water system, so you don’t have to do it a bottle at a time, or even just fill a keg at a time.

    Yes there are safety concerns with all compressed gas cylinders, do your own due diligence there, and you certainly don’t need a 50lbs for home use, but a versatile standard CO2 tank and fittings is so much better than a proprietary, expensive carbonation system.

  13. I just got an adapter and a paintball tank so it all fits within the existing Sodastream station. I don’t use it every day but it’s good to have fizzy water on demand.

  14. All in all, I’m surprised that the Sodastream markup is only 100%. ($0.18 / liter vs $0.38 / liter).

    A 5 lb or 10 lb cylinder is probably more like the right size for home use, and fits into smaller kitchens better too.

    I miss my kegerator, and not just for the home-brewed beer. I used to keep one keg filled with water, sometimes throw in some lemons or limes or whatnot. Very tasty.

    1. We started with the sodastream crystal machine the one with the fancy glass bottles and I was looking for a way refilling the 425gramm canister myself. I needed a co2 bottle with a downstem to force liquid co2 out of the canister and a speciality refill adapter that fit the sodastream with a small pressure release opening

      10 kg bottle initially cost me 140 eur plus 35 eur for the adapter and we have 3 empty sodastream gas bottles to cycle through they need to be chilled in the freezer to allow the full capacity while refilling them.

      We started this in 2016 winter and now we own a aarke carbonisator which fits my design niche quite nice and still refill co2 from the big tank… According to the tracking I am doing we saved over 400 eur in retail sodastream and competitive prices form other manufacturers by buying co2 in bulk 10 kg canisters alone also we are in Germany we drink a lot of sparkling water

  15. Using this giant cylinder strikes me as not-so-safe. I have two kegerators that last months with a 5lb cylinder. The 5lb tanks at the “welding” store are sold as “beer” gas, whatever that means. My store doesn’t sell “medical” oxygen just “welding” oxygen.

  16. Dropping this at the end because all I can say is so much wrong here. THere are a couple knowledgeable in this comment set, but a lot of WTF?

    First subthread started by Clarke: Why are people arguing here? Taking a full lungful from the tank (dry, cold, etc) is not the same as the diminishing O2 over time from a leak. At about 5%, you start to get pretty foggy. You may never feel the buildup, and by the time 10% hits, you are in the damage zone. There is a lot of literature on this. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5380556/ It is one of the many gasses I need to be concerned with professionally, and I have no reason to disagree with Dr Clarke, nor should you. A reasonable question is: How to insure that, given the size of the bottle, there can not be a sufficient buildup to be a problem in the event of a leak.

    Chris and macona: The lack of clue in this subthread is truly special. The sideline into O2 gave a bunch of people a chance to be ignorant. The difference between the various O2 types is in the bottle design and the fill method, to meet the appropriate requirements (this is for air-liquid process, where the fill is by evaporating near-pure liquid O2.) For welding and cutting, the O2 needs to be well above 99.5%, but need not be the 99.999% research grade. THe bottles are as clean as they can be, as any contamination in the essentially pure O2 at pressure can lead to unpleasantness, but the bottles are not purged unless there is a reason, so there may be minor impurities that have no effect in industrial usage, but could be harmful in medical use. USP (medical) O2 bottles are required to undergo vacuum purge prior to filling, though the required purity is only 99.5%, as long as the non-O2 components are non-hazardous (I don’t have the full standard on hand, but I would presume N2, Argon, and other inerts are non-issues. Not sure about CO2 here). O2 from concentrators is a different beast, and may be much lower purity. Aviation O2 (think military pilots, flight deck supplies for commercial, unpressurized cabins, etc. The masks for passengers use a chemical generator) has additional requirements for moisture (dry. dead dry. Gotta be dewpoint below -65C (-85F)) to prevent icing in the lines and regulators. Medical, not a real issue.

    The SDS’s may differ due to the differing requirements in spec. This covers butt. Go out of the medical side, and you no longer need to worry about some issues. Use a second source when your plant goes up in flames, you are covered (been there from the client side several times). And so on. But, barring exceptional conditions….

    There is a lot more, but this is the gist. Same O2 from the same source, but the requirements differ, handling (vacuum purge and dehydration, for example), absolute purity, and testing.

    As to the original concern of using a large CO2 bottle: Likely it is fine. The suppliers I use ship the same bottles from the same stock for beverage service and welding gas. I would hope that they are properly documented and handled to detect and prevent contamination, but I am less concerned with that since My application is welding. But it is the same gas, same stock. The actual requirement may be slightly different (99.90% for general industrial, 99.95% for USP, and 99.999 research, if I recall correctly), but in practice, it is cheaper to have every bottle meet USP. Inventory control and paperwork are not cheap. Nor is liability for sending unsuitable product.

    I have a shelf full of requirements re all of this, though these days, I rarely look at the paper versions, as most is just a google away. Some of the people here should try it sometime.

    Side item: It is worth learning where the lines are with standards, regulation, and cost. It can save your butt in a pinch

  17. Fizz is fun but carbonated drinks, even the low cal ones, have been associated with a shorter lifespan. Anyway don’t abuse me if that offends you, just do your own research and decide for yourself if the epidemiology research findings are relevant to you.

    1. Uh, no. You don’t get to make an unsubstatiated claim (“have been associated with”) and then then expect me to “do my own research”. You are the person making the claim, so if you have nothing at all to back it up, STFU. You want to know what offends me? People who make a claim, then want me to prove it for them. You even allude to “epidemiology research findings”, implying that there ARE such things, but again, provide absolutely nothing. Why did you even make this comment?

    2. SodaStream and others would like to sell you flavorings and some add their own, but many of us only drink carbonated water. Surely you’re not suggesting that that’s associated with a shorter lifespan.

  18. It’s a cool project. You could also buy adapters to refill those sodastream tanks with a siphon co2 tank.
    I belive new ones need a depressor pin to open the valve and allow filling.
    It’s safer because you can store the co2 thank outside your house but it doesn’t have the “cool factor” of showing that you have a 50lb co2 thank in your kitchen.

    You could be even cooler and order a Drago Siphon kit from argentina and have a cult item! There were made in 1965 and they are still around in use 20 years after the company stop making them.

  19. I was surprised when he told her to open tank valve all the way. I’ve always been told never to open any pressurized tank weather it a welding tank or your propane barbecue tank more than half a turn. That way if anything goes wrong you can quickly shut the tank off.

    1. It depends on the valve type. Acetylene valves are sealed only by stem packing. That is why they are so tight, but it allows a full seal when opened only 1/2 a turn. They also are designed for full rated flow at that opening. Opening too far has two potential issues: difficulty closing in an emergency and packing damage due to stem wear. In many plants, it is standard practice to bubble test C2H2 valve stems when a new bottle is attached and first opened, as it is fairly common for the stem packing to need a little takup from wear or over-opening. The fill station personnel SHOULD check, but may miss one (or just not bother, in some cases)

      High pressure inert gas valves and some more modern fuel gas valves (the modern handwheel type for B-size or large C2H2, for example) have face seats under the bonnet and require full opening to prevent leakage past the stem. The stem seal may be only an O-ring (moving, so not backed up), and will leak at high pressure, as packing tight enough to prevent it would make the valve very difficult to operate. The C2H2 may not leak even if not seated open, as the pressure is low enough that the O-ring may do, but check with the valve manufacturer.

      The general is that if it has a handwheel, full open for leak-free (but double check with acetylene to be sure)

      Once you leave hte general industrial sphere (medical O2 valves, for example) you need to know the specific valve type for proper operation.

      If you are only opening your BBQ tank 1/2 turn, that may be why it takes forever to heat. The tanks I get here specify on the label “OPD valve. Open fully”. Partial opening restricts flow and the burners may even go out.

      See, for example, the Sherwood catalog for cutaways and descriptions.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.