Here is a free guide to hemp extraction and refinement that I am working on. The main goal of this is to help make the industry a better place with better product.
If you have anything to add or want some help with your oil let me know. I also have a pdf if you want it send me an email. Its a work in progress.
Introduction
After I got my degree in Chemistry, I wanted to get into the cannabis industry in 2017 and found a book called American Herbal Pharmacopoeia which was mostly about growing but had a small section about extraction. This was my first introduction to real extraction data and helped me get a good job in hemp extraction. Hopefully someone who is getting into the industry will find this useful and maybe can get a job from the information written here like I did.
Below are my current thoughts on the processes and market, I could be wrong so take it with a grain of salt.
Pre-Processing
The pre-processing of flower is critical to any extractor as it will set up a good or bad day. It can also take a profitable deal into a net zero or negative deal and it will be hard to understand why.
A standard flower in the market is 9-12% CBD with 1-2% other cannabinoids that is free of significant stalks, seeds, dirt, debris, and is milled to spec.
Remove Stalk
The stalk of the plant will provide no CBD and will inflate the weight of the flower lot and decrease the overall potency. In some cases the flower will be tested and there will be 10% by weight of stalk that has no CBD content. In a 100,000lb lot that means 10,000lbs is useless weight and the CBD would drop from 8.0% to 7.2%. That is 363kg of pure CBD or 726kg of oil at 50% CBD.
Most growers and extractors understand that the stalk will inflate the weight and will send in samples of both stalk and flower. It did cause a lot of problems early in the industry.
Remove Seeds
The seeds will cause havoc in your oil. It will inflate the mass yield, it will take up the solubility of the ethanol or solvent, it will create an oil with low potency. The oil can be 50% of the oil weight, where a 60% CBD oil is now 30% CBD. It can be removed with distillation but it will double the input cost of distillation since there is twice the oil. It may even need to be ran twice because of the increased tails.
Typically you can get away with some seeds that have not been broken my milling but it is best to avoid it as the seeds also provide a inflated weight since they have no CBD content.
Milling
The flower mill spec will depend on the type of extraction, though most extractors prefer 0.25-0.5”. If the stalk and seeds are not removed by this point then it will be virtually impossible to remove them after.
Water Content
The water content of flower can be around 5-15% which again is fake weight that contains no CBD. Most potency analysis takes moisture content into account and will actually remove it from the potency shown. This means that someone can sell you flower with a 9% potency but what you actually receive is more like 8% and you don’t understand why.
Moisture Content can also create issues with your solvent recovery as over time it can dilute your ethanol which will reduce its solubility power. It can also create issues on CO2 extraction by freezing in the storage tank or heat exchanger to condense the CO2 from a gas to liquid. This can shut your machine down for hours.
Metal and Debris
Material that comes from the farm can have metal, bags, super sacks, etc which can create problems in your pre-processing or extraction loading system. It is best to install magnets in the auger or conveyer to limit metal parts that will shut your system down. The bags and other material mainly just needs someone to watch the material being loaded in and be able to easily stop the system to remove it.
Dirt. There can be a decent amount of dirt in the bottom of the supersacks which again will cause inflated weight with no CBD content. There isn’t really a problem in the system for dirt but there is a financial issue where you paid a premium for dirt and your extractor will be less efficient by creating less oil per hour.
Potency Variation
If you receive 100,000lbs of flower at 500lbs/sack you will have 200 super sacks to manage. Now one potency test of approx. 30g of flower does not represent this entire lot of material. The farm can vary from higher to lower potency due to many factors that honestly I am not familiar with. I have found it best to create a composite sample from each truck load of approx. 20,000lbs pulling some from each bag, mixing it, then testing that material. The QC step will help your case if you have any potency issues that can be tied to the extraction yield.
Decarboxylation of Flower
Early on there was an idea to decarb the flower prior to extraction. This would add 2 main benefits:
The CBDA would be converted into CBD so the CBD recovery would be over 90%
There would be less water in the flower so you wouldn’t have the issues created from water. You could maintain pure solvent, the water wouldnt lower the solubility and it wouldn’t freeze in CO2 extraction.
While this seems like a great value add it never seemed to make sense. When you decarb the flower it will boil the water and terpenes which is difficult to deal with. The entire building would smell of terpenes and your neighbors may complain. The water vapor can build pressure and creating a pressure rated system for this is difficult and expensive.
The decarb process would work and extraction was more efficient and easier. Though it seems easier to just deal with the water and decarb in the oil form.
Extraction
The goal of any extraction is to have a high CBD recovery of around 90% and to create a oil that is sellable or usable in the market. That is typically a 60% CBD and 65% total cannabinoid oil with no sugars or wax. Most extraction can only extract down to 1% of each cannabinoid so the spent flower will have 1% of each which is why the recovery cant be 100%. This effects lower potency flower more than higher potency since the difference is less and your recover is less.
CO2 Extraction
CO2 extraction has the name of being the most clean and creating the highest quality product, though no one really cares anymore. It creates a product with 20-30% wax which is expensive to remove. To get a recovery of 90%+ of CBD from the flower then you will have to pull high waxes.
Tip: You can use ethanol as a co-solvent to improve extraction and reduce time. Though this creates 2 problems.
You may need a C1D1 room because of the vapor
Now ethanol is in your product and you need to remove it. You have to decarb it anyway but its just a pain to deal ethanol as it cant be reused. with water/
Issues: Winterization is expensive, storage tanks can freeze because of the water in the flower if they are not drained, the regulator valves can fail if the operator is sending gas too quickly, the high pressure creates safety risks for employees.
High Pressure – high and fast
This will result in the most oil in the least amount of time. The CBD recovery should be 90% or so, though it will be very high in waxes which will lower the potency and create more issues if you have to winterize. This oil will also have a worse taste since you are picking up more chlorophyll and undesirable flavors.
Low Pressure – low and slow
I found this was the best method for extracting when you had to winterize. It was roughly 25% wax and was easy to work with on winterization. This also created an oil with a nice smell and flavor.
Remove Wax In-line – buzz word
Some people say that you can do it with out pulling waxes but its not reasonable possible. You would have to have extreme controls on the separation vessels which is not likely at large scale. Waxes are a black and white game, if you have any at all you have to winterize.
If you were to attempt this then you would need to set up the separation vessels with tight temperature and pressure control. The vessel would also likely need to be very large so there isn’t product that just shoots over into the next column. Maybe you could try pushing product into the first vessel, then allowing it to settle, then pushing it to the next but its just not efficient.
Ethanol Extraction
This is the most common type of extraction in hemp right now and there are 2 schools of thought.
Extract below -40C so there are no sugars or waxes
Extract above -40C and deal with waxes or sugar problems. This product creates a fairly cheap high CBD oil when done right.
If you are able to achieve a wax free and sugar free product then you can survive in this industry. Your product can be sold to the majority of the market, the main question is cost to produce.
Issues: Maintaining a wax free/sugar free product is the difference of making money or not. The water towers can freeze in the winter and microbes can building the water. Vacuum pumps can fail allowing pressure to building the system. Chillers can fail and the water pipes need to be metal or they can melt. Hot oil can cool post solvent recovery in the pipes and will create clogs of pure oil that are very difficult to clear. You can burn out a lot of pumps this way. The boiler, cooling system, and solvent load need to match so you are condensing what you boil fast enough. If you use liquid ring pumps then the inflow of water can break the fins or the water can be stopped from the filter because material from the water clogs it stopping vacuum.
Cold Extraction – clean the first time
This extraction is done below -40C and creates a high quality product that is wax free and sugar free. Waxes or sugars can still be present if the filtration is not done correctly and the flower re-extracts in the solvent recovery.
Hot Extraction – sugar and wax
This is done above -40C and will include winterization and a sugar removal. The sugars will create significant issues in the decarb tank by blocking the port not allowing liquid to be released or in distillation where the gear pumps will seize and will make distillation almost impossible.
Extreme Cold – high potency, hard to do
I have seen some people achieve -70C extraction and it produced an oil up to 80% CBD which was impressive. Though the chiller you need is expensive and such a low temperature is not standard in the chiller world and is hard to service. Even a -40C chiller is difficult for some.
Hexane Extraction – high wax
This was mainly done at outside and created a product that was extremely high in wax. In most cases it was 50% wax which was very difficult to winterize. It was extremely cheap to produce the crude but it was expensive to winterize and the mass loss was significant.
Decarboxylation – must have
Converted the CBDA or CBGA to CBD and CBG are included in the process of extraction. This is essentially heating the oil to 120C for 12hours or less if you have a good vacuum system. Some degradation of the cannabinoid can happen if you go too long though you need to walk that line to convert all of the acidic form. An oil with 3-10% CBDA can hurt the value of the oil because someone will have to redo what you did.
Maintaining the Acidic Form – buzz word
Some buyers need to keep the acidic form of the cannabinoid of CBDA or CBGA. This is easy in CO2 extraction as there is no heat that will convert it but you will be left with a waxy oil that customers don’t want. You can try to winterize this but you would be just doing an ethanol extraction. The problem with ethanol extraction is that you have to put it through solvent recovery where heat and vacuum will be used and will decarb the CBDA. It is difficult to remove all of the solvent and maintain all of the CBDA. You can either pull the oil early with some solvent or decarb some of the acid. While customers ask for this, it doesn’t seem to be large enough of the market to address.
Secondary Processes
These steps are generally used to get the product ready for isolation or maybe if a product needs straight distillate which isn’t very common. People seem to prefer the isolate for CBD or a converted product for some other cannabinoid. While distillation is a secondary process it is almost always used in conversion as a step to clean up the oil for solvents or color.
Winterization – must have
Waxes in this industry can be a killer for your oil as it will make a kilo valued at $120/kg be completely useless and worth $40/kg. The main test is 10:1 methanol to oil and you will check with a light if it is clear or not. Some people use a freezer to see the wax fall out but most people will accept an oil at room temp if it makes a converted product that doesn’t crash out in wax.
Most people do this with ethanol at -40C and then filtration down to 1 micron to remove the wax.
Tip: You can convert you ethanol extraction system into a winterization system if you find there is a large supply of waxy product or you have some yourself. It can be done with methanol at a higher temp.
Sugar Removal – must have
The sugar removal can be done in 2 ways
It can be heated for an extended period at 100C or quickly heated at 160C so the sugars fall out. This will create a inflated extraction mass yield as this is wasted product.
It can be mixed with heptane to dissolve the oil, then mixed with water and agitated for a period. The sugars and water soluble products will go into the water layer and then you can decant or pull from the bottom until the 2 layers are separated. There are issues with this because you may have a hazardous waste stream from the water and the heptane may not be fully recovered with water or terpenes in the heptane.
Distillation – high value add
This is needed by everyone though it is very cheap to do if you get a large enough unit that can achieve 83% total cannabinoid. Most companies have brough this inhouse because of the value add.
The exact parameters depend on flow, temperature, vacuum, and type of oil. Most people may use 120C for the first column to boil off any water or terpenes for the first “light” fraction. The second column would be 160C or so depending on the vacuum and flow to boil off the cannabinoids for the second “body” fraction leaving the heavies in the third “tails” fraction.
Sugars can cause havoc in distillation and can ruin gear pump seals and completely stop production. It is important to run sugar tests on your product prior to putting it into the system.
Tertiary Processes
These steps include isolate, T-Free, and chromatography which are very refined products.
Isolation – must have
Isolation is a critical step in the industry. It is used for consistent formulation and most conversions. At this point the product is 98-99% CBD and is generally an off white to pure white color.
To properly crash an oil into isolate you need a few characteristics:
The distillate should be over 80% CBD, it can be less but it will effect the yield and make an expensive kilo
The minors should be less than 10%, the minors can inhibit the crash of CBD and reduce yield, making a more expensive kilo
You should cool the solution to its crash point (where it starts to get cloudy), mark the temperature then heat it some and cool it again. When you are getting close to that temp you should seed the crash with pure isolate to help the crash.
The cooling is generally done on a curve or stepped cooling so the crystals do not encapsulate oil or other cannabinoids. You can do a fast crash where some oil or other cannabinoids are captured and still achieve 98% CBD.
Pentane is the preferred solvent, though some use heptane.
THC Free – buzz word
This process was extremely popular in 2020, since everyone was concerned about THC being in the oil. Though no one cares anymore since it can just be diluted to legal levels.
The two main methods here are degradation and chromatography. The degradation model is essentially just taking distillate, heating it for awhile to degrade the THC, and maybe adding isolate to reduce the THC further. The chromatography path is just taking distillate and pulling out the THC fraction.
Chromatography
Chromatography is used in many industries to pull off the desired chemical and hemp is no different. It can be used to pull out virtually any compound in the distillate, though the draw backs are that it is expensive and time consuming. There can be models where chromatography works but I think the market needs to be more advanced for there to be larger adoption.
Conversions
This part of hemp is relatively new and started with D8. The main goal here is to take CBD and organic chemistry to make the desired compound. In my opinion there are two types of conversions:
Compounds that are made with CBD as the base.
Compounds that are made with some other compound as the base.
I personally prefer the compounds that are made with CBD as the base since it helps stimulate the hemp industry instead of a chemical manufacturer. Some groups claim that these conversions are unfit for consumption though it is a bit hypocritical to allow cigarettes or vapes to be smoked and to put hemp under the magnifying glass. “You can poison yourself with the food you eat, the tobacco you smoke, and pharmaceutical pills you take but the hemp that’s where we draw the line…” These compounds should be studied to show the side reactions and the final product should be clean of catalysts, other compounds, and solvents to ensure that it is safe.
Ratings
Pre-Processing
Feasibility in Hemp Today: A
Quality of Product: A
Value Add: A
Cost of Product: Even
Market Size: 100%
Use Case: Any extractor
Facility Buildout: Partially outdoor, or indoor
Equipment: Conveyers, sorting tables, grinder, hoppers – the number one goal here is to maintain CBD content so everything needs to be in closed so the CBD doesn’t end up on the floor.
CO2 Extraction
Feasibility in Hemp Today: C
Quality of Product: A+
Value Add: C
Cost of Product: Expensive
Market Size: 5% maybe
Use Case: Organic products, tinctures, salves
Facility Buildout: Not bad, mainly CO2 alarms
Equipment: CO2 storage tank, liquid pump, extraction vessels, separation vessels, pressure regulators, chiller, heat exchanger
Ethanol Extraction
Feasibility in Hemp Today: A
Quality of Product: B
Value Add: A
Cost of Product: Controlled by isolate
Market Size: 90%, it is the base for everything
Use Case: Distillate, isolate, any conversion product
Facility Buildout: Expensive, C1D1 rooms
Equipment: Extraction centrifuge, holding tanks, filtration, solvent recovery, decarb, chiller, water tower, vacuum system, boiler
Hexane Extraction
Feasibility in Hemp Today: C
Quality of Product: C
Value Add: C
Cost of Product: Controlled by isolate
Market Size: 1%, I haven’t seen this done well
Use Case: Distillate, isolate, any conversion product
Facility Buildout: Can be expensive, C1D1 rooms, some people do this outside
Equipment: Extraction tank, holding tanks, filtration, solvent recovery, decarb, water tower, vacuum system, boiler
Winterization
Feasibility in Hemp Today: A
Quality of Product: A
Value Add: A
Cost of Product: Controlled by isolate
Market Size: 10%, most extractors remove waxes themselves though people have chillers or filters fail all the time.
Use Case: Distillate, isolate, any conversion product
Facility Buildout: Expensive, C1D1 rooms
Equipment: Mixing tank, holding tanks, large filtration, solvent recovery, decarb, water tower, vacuum system, boiler
Sugar Removal
Feasibility in Hemp Today: A
Quality of Product: A
Value Add: A
Cost of Product: Controlled by isolate
Market Size: 10%, most extractors remove waxes themselves though people have chillers or filters fail all the time.
Use Case: Distillate, isolate, any conversion product
Facility Buildout: Expensive, C1D1 rooms
Equipment: Mixing tank, holding tanks, large filtration, solvent recovery, decarb, water tower, vacuum system, boiler
Distillation
Feasibility in Hemp Today: A
Quality of Product: A
Value Add: A
Cost of Product: Controlled by isolate
Market Size: 95% virtually the entire market goes through distillation
Use Case: Isolate, any conversion product
Facility Buildout: Cheap, just a room with power
Equipment: Holding tanks, decarb/solvent strip tank, distillation unit, chillers, heaters, vacuum pumps
Quality Control
CO2 Extraction
Ensure flower is milled to a density allowing for the right pack rate. Not too fine or it will get into the oil.
Measure water content to understand the water that will collect in the oil and storage tank.
Don’t set the chiller to 0C or the water will freeze and stop flow.
Low CBD flower will make poor oil
Ethanol Extraction
Stay cold and keep it cold. While the ethanol in your storage tank shows -40C you are about to add it to a relatively hot mass that will heat up the liquid. You may need to create a recirculation look with a heat exchanger to keep it cold.
Filter down to 1 micron or less. If flower passes through the filter it will heat up in the SRU and re-extract
Measure purity of the ethanol, water or terpenes can lower the solubility power of the ethanol.
Distillation
Higher quality crude will make better distillate.
Some crude will need to go through a solvent removal step prior to distillation.
Check for sugar content and understand your units limits.
Check for waxes, they don’t cause a problem in your system but distillation wont remove waxes. You will be putting money into an oil that will have to go through winterization and then distillation again.
Formulas
Some of these will be excel formulas, the others are normal.
Flower
Convert 1,000lb to kg
Convert(1000,”lbm”,”kg”)
CBD content in flower in kg
Convert(1000,”lbm”,”kg”)*potency
Remove moisture content (get this from the 3rd party lab or do it yourself)
Convert(1000,“lbm”,“kg”) * (1 - moisture content)"
Tip: always use a calculator to convert pounds to kg, do not use 2.204 or 0.453 unless you are doing some short hand.
Extraction
Mass yield using kg flower to kg oil
(kg of oil)/(kg of flower)100%
CBD recovery of 1000kg of flower at 9% CBD and 120kg of oil at 60% CBD
(1200.60)/(10000.09)100%=83.33%
Expected mass yield – there is an issue with this formula. Typically you should be able to do a CBD recovery percentage of the extraction and then guess the potency of the oil like below
kg of CBD in flowerCBD recovery=(kg of CBD in oil)/( guessed CBD potency of oil)=kg of oil
For example: 1000kg of flower at 9% (90kg) with a 90% extraction recovery and a guessed 60% potency oil creates 135kg of oil.
90kg90%=81kg/(60%)=135kg
There are a few problems here:
It is expensive and difficult to characterize the potency of 100,000lbs of flower
The CBD recovery is rarely perfect unless you have tight controls on everything
There can be other oils in the flower that create more oil but lower potency
These other oils can compete with the CBD in the solubility of the ethanol leaving more CBD in the flower.
Sugars and waxes can create more oil but lower potency
The sugars and wax do the same as the other oil
The input weights are not always perfect with debris removed and fake weight can be added from super sacks
The expected mass yield that I prefer is
Solvent Ratio
(kg of flower)/(liter of solution)
Secondary Processing
Mass yield
(kg of input)/(kg of output)*100%
