The Whole Plant Extraction Blog

Why Proper Design of Oil Separator Vessels is Supercritical to CO2 Extraction.

Fritz Chess

The most complicated and important part of any supercritical CO2 extractor is the oil separators. This is where the extracted oil is separated from the stream of CO2 that the oil is entrained in after it gets extracted.

How Supercritcal CO2 Extraction Works

In short, the process works like this: CO2 is pumped through a pressure vessel that is packed with the material to be extracted. The CO2 is kept pressurized in a liquid, or supercritical state, by a regulating valve downstream of the extractor. Extraction pressure can be adjusted up or down by tightening or loosening the setting on the pressure regulating valve. As the stream passes through the valve, it flows into the oil separator vessel while there is a simultaneous drop in pressure. The pressure drop causes the CO2 to flash to vapor which separates it from the oil which was dissolved in it. The oil falls to the bottom of the separator while the vapor exits the top.

In older designs, the vapor leaving the separator was simply vented to the outside so the extraction system was linear meaning CO2 from a supply source was pumped through the extractor and then disposed of after it came through the separator. Most modern systems are "closed loop" in that the CO2 is recovered after the separation process and sent back to the pump for more extraction. The CO2 recovery design is integral to the oil separation process so any discussion of one must involve the other.

Liquid vs. Gas CO2 Recovery

There are two basic styles of CO2 recovery: liquid and vapor. A vapor design is fairly simple. The CO2 vapor exits the separator and enters the inlet of a gas booster pump which re-pressurizes it and directs it back through the extractor vessel. Liquid systems are more sophisticated. The vapor coming from the separator must be turned back to liquid before it enters the inlet of the pump.

There are two ways this can be accomplished: distillation and compression. Distilling CO2 is very similar to more conventional distillation techniques in that the vapor is cooled sufficiently with a condenser to liquify it. The main difference from conventional distillation is that the process is done under high pressure, 800-900 psi. Compression systems use a gas booster pump to pressurize the vapor into liquid before entering the liquid pump. These are known as dual pump systems.

Here are the pros and cons of the various pump designs:


The simplicity of the gas booster design offers great advantages from a cost and construction perspective. Separator pressure is relatively low, 350-650 psi so the separators can be built out at a lower cost with thin-walled piping. The real savings, though, are in the gas recovery phase of the extraction. There is no second pump, or distillation apparatus, to liquefy vapor, so these designs can be built quickly and cheaply.

Unfortunately, the advantages of the gas booster design extend only to the manufacturer of the extractor. There are many downsides to this design from an operator perspective. Most importantly, they are very slow. A cubic foot of liquid CO2 is 45lb whereas a cubic foot of vapor is 5lb. This means that if you have a liquid and a vapor pump of identical stroke rate and piston size, the liquid pump is moving nine times as much CO2 due to the density difference. That means nine times faster extraction!

...if you have a liquid and a vapor pump of identical stroke rate and piston size, the liquid pump is moving nine times as much CO2 due to the density difference. That means nine times faster extraction!

The low pressure of the separators and the nature of gas booster pumps presents two additional problems in terms of extract quality and pump maintenance. Most CO2 extracts contain fractions of light volatile oils known as terpenes or essential oils. Because of their low boiling points and volatility, these oils have a tendency to exit the separators along with the CO2 vapor and enter the gas booster inlet. Once they enter the pump, they are carbonized from extreme compression and heat and develop a gritty texture which damages the internals of the pump. So not only is the pump wrecked, but these precious compounds which give the oil its taste, aroma and many of its health benefits, are destroyed as well.

Another issue with gas booster systems has less to do with the basic design and more to do with cost saving measures by manufacturers and attempts by frustrated operators to make these systems go faster. Because of the low pressure that the separators operate at, some companies build the separators out of sanitary tubing and fittings which are generally rated at 650 psi. The problem is that things can go wrong which causes pressures to spike. Pressure relief valves often clog due to freezing which occurs during these events.

Consider a Nightmare Scenario

Here is an extreme instance which took place with a booster system a few years ago: In an attempt to make their booster system go faster, they added additional pumping speed. The higher flow rate caused the separator to start backing up and filling with dry ice from the rapidly decompressing CO2. This is when the pressure relief valves, a built in safety measure, clog with ice and fail. The operator then added heat to the separator to melt the ice and get the system flowing again. The rapidly melting dry ice raises pressure quickly causing vessel failure due to the clogged pressure relief valve.

It is a generally accepted axiom that any machinery that works under pressure uses a four times safety factor. This means the vessels and piping will not fail until it reaches four times the working pressure of the system. Sanitary piping does not meet this standard. Tests done by Eden Labs show that sanitary caps begin to bow at 1,100 psi. This is technically considered failure. Catastrophic failure occurs at 1,400 psi. The fact that many of these designs have received safety certifications speaks more to a flawed regulatory system and should not give anyone comfort.


The dual pumping design was developed and patented by Supercritical Fluid Technologies (SFT) in the 1980's. By the late eighties, they had rejected the design due to a lot of the same problems outlined above. Dual pumping systems tend to flow faster than conventional gas booster designs due mainly to the fact that the liquid pump raises the inlet pressure of the booster pump thus speeding the process. The main advantage to a dual pumping system is a minor savings in energy costs. Compressing CO2 into liquid with a booster pump is slightly less costly than distilling it. The big problem goes back to pump maintenance. Not only do you have the same issue with the booster pump internals being shredded by carbonized terpenes, but the liquid pump in the dual system will also need periodic maintenance. Pumps are always the weak link in any kind of mechanical system - why use a dual pump system when a single pump does the job?

Nobody understands this better than the staff of the Yakima Chief hops extraction plant. Yakima Chief is known worldwide as one of the best industrial supercritical CO2 extraction plants. It can extract six tons of hops per day. Built in 1999, the original design used the dual pump system. After several mechanical breakdowns of their booster pump, which was the size of a small house, they switched to a liquid distilling design. The system has run flawlessly ever since.

Another issue with booster and dual pump separator designs is the use of valveless expansion technology to regulate pressure in the extractor vessel. Initially conceived as a way to minimize clogging of valves, it is actually prone to clogging. The best way to regulate pressure is with a metering needle valve. Standard pressure regulators tend to break easily as they have a lot of moving parts such as diaphragms and springs. A metering needle valve simply has a needle that screws up and down creating an adjustable orifice with no delicate parts that are prone to breakage. VET is merely a small orifice fitting which regulates pressure by restricting flow. The smaller the orifice, the more pressure required to keep up the flow rate.

To change the pressure setting, you either need to take apart the separator and switch out the orifice, or change the flowrate of the pump. Both are bad options. A metering needle valve simply requires a turning of the dial. If the VET clogs, the separator needs to be disassembled. Once again, the metering needle valve simply requires a turn to open it and release the clog.


Liquid pumping offers several advantages over the designs described above: Faster extractions, less maintenance, and better quality extracts. The key to making this method work properly lies in the design of the oil separators. Manufacturers of gas booster and dual pump systems like to say that liquid extractors burn the oil during the separation process. This claim is based on the fact that the separators are heated to higher temperatures than the other two designs.

This claim can be true with some of the liquid extractors on the market. If the separator is too small, the temperature has to be kept extra high to flash off the CO2 gas rapidly. The oil is then sitting on a hot surface which turns it a darker color and changes the fragrance profile. Terpenes are damaged and oleoresins may undergo thermally induced chemical reactions.

To overcome this issue, Eden Labs separators were carefully designed to overcome this and provide the ideal conditions for oil separation from the CO2 stream. Our separators are considerably larger, and rated to higher pressures, than all others on the market. Here's why: When the oil saturated stream of CO2 passes through the pressure regulator valve and decompresses into the separator, it shoots straight down into a cool pool of liquid CO2 in our specially designed collection cup. Many other designs use a "venturi" or "cyclonic separator" system which sprays the extract against the hot sidewalls of the separator in a circular motion with the intent to enhance separation efficiently in an undersized, low pressure vessel. By using a much larger and wider vessel with higher pressure rating, Eden has created the ideal environment for harvesting oil that retains its complete constituent profile without being compromised by thermal damage.

This ideal environment is 60F and 900 psi which is the boiling temperature of liquid CO2 at 900 psi. So even though the outside of the oil separator is heated to 140F, the flow of oil saturated CO2 coming in is carefully balanced with the flow of vapor out so there is always a 60 degree pool of liquid CO2 keeping the oil cool. The higher pressure of our separators, and the cool temperatures, insures that volatile monoterpene vapors stay in solution at the bottom of the separator. The small amounts that do escape become entrained in liquid where they pass harmlessly through the pump and get another chance to end up in the separator on the next pass. These same terpene vapors would be compressed and carbonized in a booster or dual pump system.

Eden Labs Systems Produce Superior CO2 Oil

So what's the proof that Eden Labs CO2 systems make superior quality oil and extracts? Every manufacturer claims their design is the best so the real proof would come from an outside source with third-party scientific data to back it up.

Every summer, Seattle hosts the Terpestival extracts competition. Contestants submit their products which are then judged by taste and aroma as well as by modern lab analytics. The featured speaker of the 2016 and 2017 Terpsetival's was the preeminent terpene scientist, Ethan Russo, formerly of GW Pharmaceuticals.

For two years straight, all of the winning CO2 extract entries have been made with Eden Labs equipment. Heylo Cannabis and Optimum Extracts from Seattle swept the Terpestival event in 2017. These results speak for themselves. If you want to make premium quality CO2 extracts, Eden Labs is a distinguished choice. We invite you to contact our sales team to put you in touch with an Eden Labs customer near you to learn of their experience.
February 12, 2018