Open source CXE retrofit for Co2 extractors


First let me give thanks for all those who are responsible for this wonderful forum. I started learning about Cannabis on overgrow and moved to ICmag after the fallout. These forums stand the test of time and will teach the next generation of cannabis scientists. Thank you

I would like to create a open source kit to convert old supercritical extractors into CXE extractors. There are hundreds of Co2 extractors sitting idle because the owners could not justify running supercritical extractions. The bones of these extractors are good, the parameters they operate at are not.

I purpose that we create a source of components that can be used to repair the shortcomings of these old supercritical designs. With a addition of a pump and heat exchanger we could transform these old designs into a viable CXE extraction system.

I need to know is would you rather have:
A- a turn-key skid that can be added to a existing system
B- a list of components that can be purchased and assembled to retrofit a extractor.

A turn key solution would include a german triplex pump, a metered injection pump, and a brazed plate heat exchanger. These additions engineered on a skid would cost around $9k.

The alternative would be to keeping the OEM pumping and condensing units and add a metered injection pump. This option would cost around $3k and allow for CXE but not address the pump and condensing problems that plague supercritical Co2 machines.

Please let me know what you think and add any ideas that you may have. I consider this a community project and if we can save just one Co2 extractor from collecting dust then I would consider it a success.


my first response involved language I’d rather not leave lying around. :slight_smile:

lets just call it enthusiastic…

Open source? same unprintable enthusiasm.

$9k for a skidded add on seems very reasonable. If it’s truly open, I assume the other options are not out of the question, just not first out the door.


Hi Jon,

This is a great idea, thank you for considering open source. In the meantime, how necessary is it for the EtOH to be pumped in to the system? I’ve been wanting to try extracting the lighter fractions with CO2, then just opening the extractor to add EtOH to get all the stubborn THCA. I’m sure it isnt perfect, but anything to raise the yield/hour problem. Am I missing something here?

Do you use any kind of static mixer to get the CO2 and EtOH dissolved, or do they mix pretty easily?

Is the heat exchanger to cool the ethanol before it goes in?

You can do a “pour over” with ethanol and then rinse with Co2 and it will increase the speed of extraction. The extract that results from this technique will have similar non target compounds as a ethanol only extract. For most this will be sufficient because post processing can remove non targets.

When you inject the ethanol using a metered pump you can dial in a percentage charge and keep that ratio throughout the cycle. As the Co2 extractor condenses pure lCo2 the metered injection charges the fresh solvent with modifier. This technique can avoid the non targets while increasing the extraction speed and allowing for a winterization filter right out of the extractor (so long as the separator temps are not high).

The Co2 and ethanol mix readily so long as the charge is not above 10%.

The heat exchanger is to increase the condensing power to keep the triplex pump flooded with liquid. If we hooked up the triplex pump to a tube in tube heat exchanger the pump would run on 5% power max before running dry. We want to over power the extraction system with condensing so we can keep the pressure and temperature low.


Thanks for the thorough response!

That is really interesting that the heat exchanger might be the limiting factor on these systems. I always figured it was how much people were willing to spend on high pressure pumps and chillers.

How do you get your lCO2 flow rate to meter your modifier? Ideal flow rate based on pressure curves of the pump?

I’m assuming you flow top to bottom through the extractor so the EtOH drains out the bottom? Does the EtOH all stay in the separator or does some build up in the accumulator or get recycled through the pump you think?

The Co2 flow rate is measured by sight glasses and the time it takes to fill a vessel with liquid Co2. You can use float meters also but a sight glass allows you to see the color of the solvent mixture which really helps to dial in the ratio of cosolvent addition.

The cosolvent will all end up in the separator. Ideally you want a drain in the collection cup to remove the cosolvent mixture throughout the cycle. Without a drain you will have to stop the cycle and vent the separators to drain the cosolvent out of the machine.


Ooh a sight glass sounds so nice! Also loving the idea of pulling extract out in liquid EtOH rather than a sticky glob of baby shit.

I’m really liking the idea of retrofitting these systems to make them usable, and the float meter sounds perfect, cheap and simple. Anything to let the operator know what is going on in the system other than trial and error.

Speaking of floats, any ideas for a level sensor in the separator. I think a lot of people have trouble dialing in that liquid level to not flood or completely dry out and cook the extract.

Also interested in adding a UV-vis or some kind of detector to better resolve fractions without more trial and error. Maybe this could help with the EtOH/CO2 ratio as well.

Anyways, thanks for the info man, you are extremely helpful, and definitely helped me get terpene extractions going back in the day. I still believe CO2 is a viable extraction method, especially when flavor is a priority, and CxE might just be what makes it competitive.

Just reread your post. To answer your origInal question, the cheaper option of just the solvent injection pump is probably easier to sell to the people with the $. The full upgrade is very interesting. What size system and flow rates are you talkin about for those prices?

The full upgrade would only make sense for extractors larger than 10 liters.

A eden will be able to flow up to 10 liters a minute because of the accumulator.

A apeks will be around 8 liters a minute because it has no liquid storage.

The triplex pump can flow way more than 10l/min but both the Eden and Apeks are restriction type designs and the needle valve/ nozzle can only flow so much into the separators. That is the real bottle neck to these supercritical extractors.


Thank you for all you do sir. you have provided a beacon hope to many. I suggest making turnkey skid and open source info for DIY before Apeks tries to make out like its thier idea LOL


This. They are notorious for their patent on the “Valveless” expansion technology and I would hate to see a good community idea be sequestered away as “IP” or something akin.

Just a thought, if you are using ethanol only not using it as a co-solvent with co2, could you remove the Micron filter in the Apeks separator and/or buy a piece of tubing to replace/bypass the needle valve on Eden and not have the flow rate problem? I may be missing something though because this solution sounds too simple.

Sidewinder pumps. Call and talk to Guy about how much you want to flow per day (it’s weird but just figure up your ideal injection to his per day standard) so he can size your pump. Let him know the pressure you want to inject at and he will get you set.

You have the air driven options of the 120v injectors. Both are straightforward and easy to use. Modifier on the inlet check valve on the outlet. You want to tube the pump outlet to your machine with a tee and another check valve. Ideally you want to inject after the pump and before the extraction vessel.

1 Like

Will this work for IES systems? 50L and 100L?