Processors have a tight moisture target to hit when drying cannabis. Overdrying damages trichomes and affects yield. Under-drying violates regulations, increases the chances of mold contamination, and puts customers at risk.
This guide explains the two primary ways of measuring moisture in a porous medium like cannabis and looks at how, when, and why cannabis processors need to monitor and control moisture drying cannabis. This control should extend into packaging, storing, and shipping.
Water activity—a critical way to evaluate moisture
Cannabis processors are already comfortable with the concept of water activity (aw)—they just know it as relative humidity (RH). Relative humidity is measured in air and expressed on a scale from 0-100%. Water activity is the same concept applied to a porous medium like food, pharmaceuticals, or cannabis, and it’s expressed on a scale from 0-1.00. When cannabis flower is in equilibrium with the air around it, the relative humidity of the air will be equal to the water activity of the cannabis x 100 and expressed as a percentage.
When to stop drying cannabis
Relative humidity is already a consideration at most cannabis processing operations. The bud is put into drying rooms at a specific relative humidity and dried. But how do you know when to end drying?
It’s a bit like cooking meat in an oven. Turkey, for example, is typically roasted at 325 °F, but it’s not being cooked to 325 °F. Turkey with an internal temperature of 325°F degrees would be inedible. You know when the turkey is done by measuring the temperature of the turkey itself.
In order to do that, you need to:
Have a definition of “done”
Safety defines the lower limit. The USDA recommends that turkey be cooked to a minimum of 165°F.
Quality defines the upper limit. It varies depending on the preferences of the cook and the type of meat and can be anywhere between 165 °F for breast meat and 180 °F for thigh meat.
Using the safety and quality boundary numbers, you can define a “sweet spot” where your meat will be safe and tasty.
Use the correct measurement to monitor condition
When cooking a turkey, you need an accurate and properly calibrated instant-read thermometer. You also need to know how to use it correctly to measure temperature during cooking. Other frequently used methods, such as setting a kitchen timer, watching to see when the juices run clear, or testing looseness of the thigh bone will not give consistent, reliable results.
Characterize your process
As experienced cooks will tell you, the temperature of the turkey will rise by a few degrees after it’s out of the oven, so you will need to take that into account.
Like temperature, water activity/RH is an intensive variable that can help you determine doneness. And just like in the turkey example, there are three things you need:
1. A water activity definition of “done”
Safety defines the upper limit as 0.65 aw. Any product at water activities higher than 0.65 aw is susceptible to mold growth.
Quality defines the lower limit as 0.55 aw. This is the generally accepted value. Below 0.55, terpenes dry up and dried cannabis becomes unacceptable from a quality perspective.
This puts the sweet spot for cannabis somewhere between 0.55-0.65 aw.
2. An accurate water activity meter
When you drying cannabis flower, you have to get the entire crop to the perfect moisture level. There are many factors involved, including the size of your batch, the relative humidity of your dry room, whether you dry on trays or wires, and airflow patterns during the drying cycle. The only way to evaluate and improve your drying methods is to measure the water activity/RH of different samples in different locations during the drying process.
3. A well-characterized cannabis drying process
Water activities will inevitably vary across locations in the drying room, but if you use a curing or burping step, the entire batch will equilibrate to a single water activity. Predictive algorithms can help you make decisions at each step that will result in a crop that hits the sweet spot of safety and maximum yield/quality after burping every time.
If you attempt to measure doneness using time, subjective tests such as the snap test, or less-relevant moisture measurements such as moisture content, the results will be inconsistent. When you base cannabis drying on a high accuracy water activity measurement, you can develop analytics that will give you consistent, safe, high-quality results from batch to batch. Measuring water activity is easy. The AQUALAB 4TE provides an accurate aw number in as little as five minutes. Watch the video to see how it works.
Drying cannabis to control mold
Water activity controls mold and microbial growth. That’s the scientific fact that makes water activity a key part of food and pharmaceutical regulations. Different species of mold (and microbes) have specific water activity limits below which they will not grow. Watch the video to see why.
Water is essential for all living organisms, including bacteria and mold. Microbial cells need it to remain viable. Cells bring in water via passive transport. When the water activity gets low enough outside the cell, the cells can’t transport water in through the cell membrane. They starve and go dormant.
It’s important to note that these cells are not dead. if the water activity outside the cell goes above the growth limit, they can start to grow again. However, cannabis can be processed, packaged, and stored below water activity limits to ensure that microbes and molds will not grow on or in it. Learn more about water activity and microbial growth in the following 30-minute webinar. Learn:
- what you need to know about how water activity predicts microbial growth
- how to use specific organism aw limits relevant to cannabis in setting your specs
Cannabis and moisture content
Water activity is the correct measure to use when predicting whether or not mold and microbes will grow in or on dried cannabis flower, but it doesn’t give you any information about yield.
In fact, water activity is an intensive variable. Measurements like water activity, relative humidity, temperature, and pressure describe the intensity of something, not the amount. That’s why they’re called intensive variables.
In contrast, variables that depend on the amount of matter being measured, like moisture content, energy, force, time, and area, are called extensive variables.
To illustrate the difference, think about how the intensive and extensive variables work when heating your home. Temperature, the intensive variable, is the correct value to use in determining how to heat your home. It describes quality–how it feels to be inside the house–and safety–whether the temperature is adequate for the living organisms in the house, from plants to pets to people. It would be difficult to control your thermostat correctly by monitoring and controlling an extensive variable like number of kilowatt hours or volume of natural gas.
When it comes time to pay, amount becomes the key factor. The temperature on the thermostat can’t tell you anything about how much heat you used, or what your bill will be.
Similarly, the intensive variable, water activity, should be used to control the cannabis drying process while the extensive variable, moisture content, should be used to measure yield. Regulatory agencies also typically need to know moisture content to make sure you are not over-diluting the potency of your cannabis product.
Measuring moisture content: speed vs. “accuracy”
There are many different methods of determining moisture content in cannabis, but the most common is loss-on drying. The sample is weighed, dried, and weighed again.
Assuming you’ve gotten the cannabis sample fully dry, you can take the difference in weight and determine what percentage of the sample was water (for more details on this, see the “Scientific definition of moisture content” section of this guide).
There are several assumptions behind this method. The first is the assumption that the sample is fully dry. But there is no way to know whether or not you have removed all the moisture from your cannabis sample. As a result, there’s no definition of “dry” for any substance, and no standard to use in evaluating different methods of loss-on-drying. That’s why the word accuracy in the title of this section is in quotation marks. Because it is not possible to evaluate measurement methods against an accepted standard, it is not possible to determine the accuracy of each method. When you see a stated accuracy in relation to a moisture content method, it is more likely referring to precision.
The second assumption is that everything driven off when you dry the sample is water.
In 2015, Dr. Aaron Stancik of CannaSafe Labs tested this assumption by drying cannabis flower in several different ways. He used a very low-temperature vacuum oven (35 °C), periodically testing the change in weight until no further weight loss occurred (this took 2 to 3 days).
He also tested samples in a commercial instrument that uses a combination of low temperature and desiccated air to dry the sample and automatically detects the point when weight stops changing. He then compared these results with those obtained using a halogen quartz fast loss-on-drying moisture balance. Results from the vacuum oven and the commercial low-heat drying instrument agreed within ±0.1% VWC. Moisture content measurements from both methods averaged about 3% less than those from the fast loss-on-drying moisture analyzer, suggesting that the fast moisture balance overstates moisture content significantly.
Dr. Stancik postulated that the fast loss-on-drying moisture balance overestimates moisture content due to the volatilization of compounds other than water. Because volatiles make up between 1 to 8% of any cannabis sample, the effects of volatilization cannot be considered insignificant. Not all cannabinoids volatilize at low temperatures, but flavonoids and terpenes certainly do.
In dried bud, fast loss-on-drying moisture balances will almost certainly deliver unstable and artificially high readings. This is of particular concern to the processor who is trying to maximize quality and yield in the cannabis crop while avoiding the risk of mold growth. Dr. Stancik’s low-temperature method will deliver reliable readings in two to three days, but for those who need faster measurements, we will discuss some alternatives in the next section.
Alternatives to loss-on-drying cannabis
Karl Fischer titration is commonly used in the pharmaceutical industry to determine water content. The method is based on the Bunsen Reaction, where water, sulfur dioxide, and iodine react to form sulfuric acid and hydrogen iodide. Because the reaction consumes water, German researcher Karl Fischer realized that the amount of water in a sample could be calculated based on the concentration of iodine remaining after all water has been consumed. The method can be very precise, but it is expensive, requires expertise, and uses hazardous chemical reagents.
Moisture sorption isotherms offer an alternative way to measure moisture in the cannabis drying process. This method allows you to determine moisture content in less than five minutes and can be very precise. The method uses a moisture sorption isotherm, a graph that plots how water activity changes as the dried cannabis flower absorbs and desorbs water.
The isotherm is created by drying cannabis flower all the way down and measuring how the moisture content and water activity change during the drying process. This relationship between water activity and moisture content is different for every product, but a specific product’s isotherm can be used to determine that product’s moisture content by measuring its water activity.
Isotherms for a number of different strains of dried cannabis show that cannabis has a very consistent isotherm. This makes it possible to measure water activity on a dried cannabis sample and determine the corresponding moisture content value with a high degree of precision. In testing, this method has proven to be as precise as a low-temperature vacuum oven in determining the moisture content of cannabis.
There are commercial instruments that store isotherm models for specific products (see AQUALAB VSA). These instruments display both moisture content and water activity for a dried cannabis sample. This makes it possible to do a single test in under five minutes that gives both a highly accurate water activity value and a precise moisture content measurement.
Currently, low-temperature vacuum oven loss-on-drying and moisture sorption-based water activity methods are the best methods for most processors to use in determining moisture content in cannabis flower.
The moisture model: how to maximize quality and yield
Inaccuracy can have a significant impact on yield. For example, take a processor drying cannabis to 11% moisture. Most processors are using inexpensive loss-on-drying instruments, which have at best ±1% precision. When drying to 11% moisture, the actual variation in the moisture of the dried cannabis will fall between 10 to 12%
This generic isotherm for cannabis flower is a moisture model that shows how water activity changes as you dry the cannabis. It is quite stable and applies with minor offsets to most varieties of cannabis.
The y axis shows the full range of moisture content and the x axis shows what the water activity of cannabis is as the bud is being dried.
Essentially, the y axis of this graph tells you about the quantity of water in your product—which most people think about as yield—and the x axis tells you about quality or water activity.
With a moisture content of 11 ±1%, this processor is anywhere between 0.53 aw, which is too dry from a quality standpoint, and 0.66 aw, which is wet enough to grow mold. That’s a huge amount of variability in the quality of the cannabis product.
It could be beneficial for this organization to measure water activity instead and use that number to get moisture values. Here’s how that would work. Set a specification of 0.62 aw, well within the range of good quality and still on the safe side of the mold growth threshold for dried cannabis. Use a chilled-mirror dew point instrument to measure water activity. The error bars of ± 0.003 aw would put the reading between 0.617 and 0.623 aw.That translates to a moisture content between 11.16 – 11.28%, a range of 0.12% moisture content.
By measuring on the quality side, this processor:
- Gets visibility into quality—they know for sure that they won’t cross the mold limit of 0.65 aw for dried cannabis
- Eliminates issues with volatilization, because they don’t heat the cannabis sample
- Get almost 10x better precision on the yield side, improving from ±1% to ± 0.12%.
Here’s what that means for profit.
One processor calculated that a lack of precision in his moisture measurements was costing 2% annually in yield. At 125,000 lb production annually, a 2% improvement in yield represents an additional 2,500 lb production. At an average wholesale price per pound of $1000, the additional production will add 2.5 million dollars in revenue while actually reducing production costs (because the cannabis drying time is shorter).
Moisture models: details and considerations
A moisture model is simply a graph of how water activity changes as a product is dried and as it absorbs water.
This is the generic isotherm for drying cannabis. Different strains of cannabis can have widely divergent flower sizes, densities within the flower, and chemical makeup of volatiles in the dried crop. Nonetheless, testing done several years ago seems to indicate that the isotherm curves are incredibly consistent regardless of strain.
This multi-point isotherm was created by a vapor sorption analyzer instrument. Here’s how it works: A sample of cannabis flower is placed in the instrument and dried gradually by exposing it to desiccated air. As the sample is dried, the instrument stops frequently, waits until the sample weight stops changing, then takes a snapshot of the moisture content and the water activity. Each of these snapshots forms a point on the desorption curve. Once this curve has been completed, the vapor sorption analyzer begins wetting the product up with humidified air to create an adsorption curve.
Why are adsorption and desorption curves different? This is a well-known phenomenon called hysteresis. Simply put, the water-holding capacity of any material is different depending on whether it is adsorbing or desorbing water. For practical purposes, that means that the model you use to relate water activity and moisture content will depend on whether you are drying cannabis or allowing it to wet back up as it is exposed to higher humidities. The drying model applies to the cannabis drying process; the adsorption model applies to product as it is exposed to ambient humidities higher than the humidity of the product in the package. Specific working models can be designed to fit your needs.
Water activity measurement method comparison
The best way to compare water activity instruments is by sensor type. The type of sensor is typically the limiting factor when it comes to speed, accuracy, longevity, and reliability of your instrument.
In testing, we found that chilled mirror dew point sensors were five times faster on average than resistive electrolytic and capacitance sensors when measuring product samples. Accuracy, as measured on both saturated and unsaturated salt standards of known water activity value, was also significantly better for chilled mirror sensors. Based on these data, we would recommend the chilled mirror sensor as the best option for measuring water activity in dried cannabis flower.
There are several commonly-expressed concerns about chilled mirror sensors. We found only one of these to have merit. The accuracy of chilled mirror sensors can indeed be affected by samples containing certain volatile compounds in high concentrations. Testing has shown that dried cannabis flower does not contain these compounds in high enough concentrations to affect the mirror. (Note that cannabis products such as alcohol-based tinctures may contain ingredients that contain high levels of these volatile compounds. These products should be tested with a chilled mirror sensor to determine whether or not the volatiles affect the mirror.)
Other concerns, such as the belief that the sensor is more easily contaminated than other water activity sensors have proven unfounded. In recent comprehensive comparison testing, it has been shown that the chilled mirror dew point sensor is less affected by contaminates than resistive electrolytic and capacitance sensors. Find a complete evaluation of all sensors along with testing data in this article.
The chilled mirror sensor’s longevity, stability, reliability, and low cost of ownership make it the best choice for cannabis processors.
The costs of (in)accuracy
Small cultivators may find the cost of all serious water activity meters high. There are extremely low-cost options (starting at about $600) offered on major e-commerce sites. We acquired and tested a couple of these.
Our small hand-held water activity meter was labeled AMTAST. Similar meters appear to be available under the names VTSYIQI and CNYST. We also purchased the CGOLDENWALL Smart Water Activity Meter for about $1000.
The hand-held meter claims to be accurate to 0.02. We followed the instructions to put a sample in the provided cup and place the hand-held meter over the cup. We found that the instrument did not seal to the cup, and that the instrument was typically measuring ambient humidity more than sample water activity. For many samples, readings would continue to drift with no endpoint. The instrument uses time instead of reading stability to end a reading. We found performance to be unacceptable for any application.
The CGOLDENWALL came with only a single test cup. We used commercially available salt standards to test the instrument, but mid-way through the first week of testing, the instrument’s own power regulator failed and burned up a component on the probe motherboard, leaving scorch marks on the lab bench and a fried electronics smell that lingers in the lab to this day.
Our attempts to get the instrument repaired have been a cautionary tale about the costs of buying sophisticated equipment through a third-party site with no easy access to warranty support. We have finally abandoned this first instrument and acquired a second instrument to test. We will publish performance testing data as soon as it is available.
So if you’re very small and price sensitive, what are your options?
- Take advantage of 3rd party testing labs. Offsite testing is a pain and an expense, but your brand and business reputation are tied to these test results. Getting reliable results from an outside lab may prove to be less costly than a short-lived and inaccurate instrument.
- Try instrument rental and lease-to-own programs. Some of the big instrument manufacturers offer special programs for smaller businesses.
Water activity in cannabis drying standards and regulations
Each state sets its own regulations and testing requirements for dried cannabis, which can frequently change. Some states require water activity testing. Some do not. What doesn’t change from state to state are the water activity microbial growth limits. Whether or not specific states require water activity or mold testing, the quality of the dried cannabis product depends on the crop being consistently dried to a safe level.
To find out whether your state requires you to report water activity readings, how they can be taken and verified, and how they should be reported, consult your state marijuana board.
There are standards that relate either directly or indirectly to the testing of dried cannabis. These standards have been established by standards bodies, public or private organizations that have to establish, publish, and coordinate voluntary standards, guidelines that help clarify and control processes of importance to the public. As the cannabis industry develops, voluntary standards will help guide the development of regulations and can help give processors, wholesalers, manufacturers, dispensaries, retailers, and ultimately consumers the ability to compare and have confidence in the quality and safety of cannabis products. Using these standards is also a way for processors to differentiate their product in the marketplace.
Some of the most valuable of the standards are:
- ASTM D8196: Determination of Water Activity in Cannabis Flower, and
- ASTM D8197: Standard Specification for Maintaining Acceptable Water Activity Range (0.55 to 0.65) for Dry Cannabis Flower
The ASTM standards are the first to establish specific recommendations and ranges for measuring moisture in dried cannabis flower. ASTM D8196 lays out standard measurement procedures and guidelines. ASTM D8197 establishes the recommended range to address safety and quality issues.
USP Method 1112
The United States Pharmacopeia (USP) is an independent, not-for-profit, non-governmental body that sets quality, purity, strength, and identity standards for medicines, food ingredients, and dietary supplements. Many USP standards are enforceable by the FDA. Method <1112> provides “guidance in the influence of water activity as it pertains to product formulation susceptibility to microbial contamination.” The chapter includes a water activity limits microbial growth table, strategies for microbiological testing based on water activity, and methods for measuring water activity. USP <1112> is often used to justify reducing the amount of microbial limits testing needed for products with lower water activities.
This standard is issued by the Association of Official Analytical Chemists. Though it pertains officially to the determination of water activity in “canned vegetables,” section B on instruments and systems discusses acceptable methods for the determination of water activity and specifies acceptable sensors.
Compendium of Microbiology Water Activity Method Chapter 68
This chapter in the Compendium of Microbiology defines water activity, gives a brief review of measurement methods, discusses calibration and sample preparation, and includes a detailed table on microbial growth limits.
ISO 18787:2017: Determination of water activity
This standard is issued by the International Standards Organization. It “gives basic principles and requirements for physical methods of determining the water activity of products intended for human consumption and the feeding of animals.”
Health Canada Compendium of Analytical Methods MFLP-66: Laboratory Procedure for the Determination of Water Activity Using the Decagon Aqualab
Method for measuring water activity in food and food ingredients to determine compliance with the requirements of Sections 4 and 7 of Canada’s Food and Drugs Act.
A scientific definition of water activity
Water activity is a measure of the energy state of water in a system. The system might be a piece of cake, a bag of dog kibble, or a sample of dried cannabis flower.
On one end of the energy spectrum is pure, free water, unaffected by any chemical or physical bonds.
The energy of this water can be reduced in several ways. Imagine dipping a sponge into the water. The free water will move into the sponge—water always moves from high energy (high water activity) toward lower energy.
Within the sponge, the water is bound by hydrogen bonds, capillary forces, and van der Waals-London forces. These are called matric effects. The water in the sponge has a lower energy state than the pure, free water. One way of knowing this is by realizing that in order to return the water to its free state, we would have to input work by squeezing the sponge.
Water in the sponge has a lower vapor pressure, lower freezing point, and higher boiling point than pure, free water. We can measure and quantify these differences. The way we measure and describe them is through water activity.
Note that water’s energy (water activity) can also be decreased by diluting it with solutes. Just like with the sponge, work would be required to remove the solutes and return the water to its free state.
Water activity is a sum of matric and osmotic effects.
Water activity is a unitless measurement. The water activity is defined as 1.00 aw. It is important to note that because the full range of water activities are compacted into the space between 0 and 1.0, changes out to the third decimal place can be important.
Water activity is a function of temperature. The temperature of the sample should be monitored and accounted for in all water activity measurements. For the most accurate measurements, temperature must be well controlled.
A scientific definition of moisture content
Moisture content (or water content) ]is defined as the percentage of a product’s weight that is composed of water. It can be reported on either a wet basis, where the amount of water is divided by the total weight of the sample (solids plus moisture), or on a dry basis, where the amount of water is divided by the dry weight of the sample (solids only). The basis should always be specified.
The AOAC lists 35 different methods for measuring moisture content. These methods are classified as direct and indirect.
Direct methods provide the most reliable results and include oven drying, vacuum oven drying, freeze drying, distillation, Karl Fischer titration, thermogravimetric analysis, chemical desiccation, and gas chromatography.
Indirect methods measure some property of the sample that changes as moisture content changes. Indirect methods must be calibrated to a direct method. They are less precise than direct methods, because the uncertainty inherent in the calibration must be added to the uncertainty inherent in the direct method. Indirect methods include fast loss-on-drying moisture balances, NIR absorption, IR absorption, microwave adsorption, dielectric capacitance, and refractometry.
Association Official Analytical Chemists . Official Methods of Analysis of AOAC International. 21th ed. Volume II AOAC Scientific Publications; Arlington, VA, USA: 2019.
Anthony J. Fontana, August 2015: 68. Measurement of Water Activity, Acidity, and Brix
Compendium of Methods for the Microbiological Examination of Foods. https://doi.org/10.2105/MBEF.0222.073
Water Activity: Theory and Applications to Food, L.R. Rockland and L.R. Beuchat, Eds. (Marcel Dekker, New York, 1987).
A.M. Cundell, and A.J. Fontana, “Appendix D” in Water Activity Applications in the Pharmaceutical Industry, (DHII/PDA Books, Scottsdale, Arizona, 2009).