In our bi-weekly poll on the BH Facebook-group you asked us, “Can you please summarise Cafe Imports’ Water Activity Paper?”, which is based on years of cupping and data logging by the influential green coffee importers. And we said, “Yes we can.” Here is this week’s offering from BH Learn.
The importance of moisture in green coffee is well-known. Improper drying causes all kinds of defects, from mould found in coffee that isn’t dried enough, to poor storage qualities for coffees that are dried unevenly. However, to measure and monitor these effects, green coffee professionals are moving away from just measuring moisture content as a percentage, towards a measure called ‘Water Activity’ (also written as Aw).
Water activity is a relatively new measurement in speciality coffee, though it is well established in other parts of the food industry. Rather than measuring the total amount of moisture in a sample, it measures how tightly bound the water is in the sample. In a sample with a high water activity, the water is less tightly bound, and can therefore take place in chemical reactions or other processes more easily. In other industries, it is water activity rather than moisture content that has been shown to affect, for example, how quickly food goes mouldy or stale.
Since 2012, the team at Cafe Imports have been measuring the water activity of all samples sent to them – a total of 25,000 samples. They used this unusually large dataset to try and investigate some of the effects of water activity in speciality grade coffee, and to help understand whether it’s useful for farmers, green importers, and roasters to start measuring water activity more widely.
What is Water Activity?
The strict definition of water activity is: the ratio of the vapour pressure of water in a substance relative to the vapour pressure of pure water under standard conditions.
To understand vapour pressure, imagine pure water in a half-filled, sealed container. The air above the water is at atmospheric pressure (just over 1bar). However, because some of the water has evaporated into the air, some of the pressure in that air is due to water vapour, while the rest of the pressure is due to other air molecules. In fact, in this system, about 0.03 bars of pressure is down to water vapour; this is the partial vapour pressure of pure water.
Now imagine the same system, but with salty water in the bottle instead of pure water. Salty water has a lower water activity because the salt ions bind to the water, so it releases less water into the air. The partial vapour pressure of water in this system is therefore lower. The water activity of this salty water is defined as the ratio between the two — so the less water is available to the atmosphere, the lower the water activity.
Pure water thus has a water activity of 1 (a 1:1 ratio), while a substance that generates half as much vapour pressure will have a water activity of 0.5. This number is a measure of how much water is available to take part in reactions – so a food with a low water activity will not go mouldy, because the water in it can’t be used by microorganisms.
What did they find?
The team recorded Aw, moisture content, and cupping score of 25,000 samples at different stages of the buying process — pre-shipment samples sent from origin; samples taken on arrival; and spot samples taken from coffees in the warehouse at 3-month intervals. These samples covered a wide range of coffees, which scored in the 70-90 points range, and excluded decafs, ‘monsooned’ coffees, and robustas.
Water Activity vs. Moisture Content
The first important result is that the Aw of these samples very rarely fell outside accepted guidelines. The average Aw for their samples was only 0.554. Microbial growth doesn’t occur below 0.6, and the majority of the coffees fall below this point. The current upper limit for Aw in speciality coffee is 0.7, and 95% of samples submitted were reaching this standard. “This is so lax as to be meaningless,” the authors point out.
Access Our Online Courses
And Premium Content Now!
With a monthly subscription to BH Education, you’ll get access to
12 online courses, 600+ lessons, 200+ videos, powerful apps, online
calculators and lots more, to make online learning easy.
They also found that moisture content, which is considerably cheaper to measure, was “neither a perfect nor a terrible predictor of water activity.” As a result, they suggest that using a cutoff point of 11% moisture could be sufficient to minimise issues related to water activity. “In green coffee processed for specialty export it is uncommon to find 11% moisture coupled with Aw greater than 0.610,” they note.
They also took measurements of samples during drying in Costa Rica and Colombia. They found that while moisture content decreases steadily during drying, the Aw was very variable. Because of this, they conclude that when considering a sample for approval, “water activity is probably less well-suited [than moisture content] to making close calls.”
Browning in Storage and Roasting
Maillard reactions, more familiar to us from roasting coffee, can also take place slowly during storage in coffees with higher Aw. The team ran a series of experiments designed to see if this had any effect on browning in roasted coffee, using samples of coffees with different Aw, then by changing the Aw of a single coffee by using desiccants or a humid environment in storage.
In all cases, the coffees with higher Aw tended to end up lighter, suggesting that higher water activity isn’t contributing to browning during roasting. They also checked the Aw levels of samples taken at different temperatures during the Maillard phase of a roast, and found they varied considerably relative to the starting Aw, but didn’t find any effect on roast colour.
One of the problems associated with high Aw, lipid oxidation is associated with flavours such as paper, wet cardboard, or cedar wood. By analysing their cupping data, they found the highest occurrence of such flavours in coffees with Aw above 0.610. They suggest further research measuring lipid oxidation directly would be worthwhile.
To see if Aw helps to predict how well coffee will last in storage, they compared cupping scores on samples of the same coffees, taken at different stages in the buying process — from pre-shipment samples, to arrival, to spot samples. They found that coffees with higher Aw deteriorated more quickly on average at each stage, and this effect was the same for coffees from many different origins.
The coffees with the highest score at the PSS stage seem to deteriorate the most, and show the biggest effect of high Aw: coffees with Aw above 0.610 that scored 90+ at PSS dropped to an average of about 81 by the time they’ve been on spot for 3 or 4 months (compared to a range of 84 to 86 for other Aw levels).
They also found that as the coffee passed through the process, the Aw of the different coffees tended to converge as they reached equilibrium with the conditions in the warehouse. But the team point out that changes due to Aw, such as lipid oxidation, are not reversible — so the Aw need to be controlled at every stage in shipping and storage to minimise these effects.
Considering the close relationship between moisture content and Aw, the team argue that “as it stands right now, there is no urgent need to adopt water activity across the industry.” Relying on Aw in contracts and buying decisions “would impose a poorly understood and expensive measurement … on people who cannot afford it,” they say.
Instead, they suggest that Aw data would be better used to refine standards for moisture content, which is cheaper and easier to measure. From their data, they found that the commonly used upper limit of 12% moisture allows through some coffees with Aw above 0.610 – high enough to have a measurable effect on quality. They suggest that setting 11.5 or 11% as a cutoff would minimise such problems – but note that as it stands, few coffees presented to them reach this level anyway.
They conclude that current best practices — buying coffee below 11% moisture content, and then storing it in relatively cool and dry conditions — will help prevent spoilage via non-enzymatic browning and lipid oxidation, by keeping water activity below 0.6 at all stages in the process. Being able to measure Aw has some benefits to a large importer like Cafe Imports, who use it as just one part of their quality assessment programme — but other stakeholders such as producers and roasters may not consider it necessary. “The establishment of any industry standard for water activity in specialty green coffee at this time is premature,” they conclude.
(The above photo featuring a water activity meter is published with the kind permission of Cafe Imports.)