At Swiss Water, we do more than just take the caffeine out of coffee. Part of our mandate for continuous quality improvement includes constant experimentation and discovery. Our Director of Coffee, Mike Strumpf, has been conducting research in the area of water activity and green coffee sample preparation—factors that affect how you roast coffee—especially decaf. We were delighted when our importer partners in the U.K., D.R. Wakefield, invited Mike to talk about his research to an at-capacity audience of London coffee industry professionals earlier this month. Here are some highlights of the knowledge Mike shared that day, along with full video of the talk itself!
Unless you live completely off the grid, your life relies on data: specifically accurate data. Sometimes erroneous data is very obvious, like a phone number with only 6 digits, or getting search results for “Columbia” instead of “Colombia.” But sometimes, erroneous data is harder to notice.
Erroneous data in coffee can come from many sources; maybe your shot timer didn’t start right away, you didn’t tare your scale, or your thermocouple probe offset is wrong. Another source of bad data can be from your lab equipment such as moisture or water activity meters. We think about mise en place when cooking or working on bar, but do you have mise en place for lab work? Consistent preparation of the temperature and humidity of green coffee samples is of paramount importance if you want consistent and reliable results. (And you need reliable results if you want to make good buying and roasting decisions based on your data.)
Things are going to get a little bit technical here, but stay with me.
Water activity in a coffee bean is the measurement of the state of energy of water in the bean, which is different from the moisture content in the bean—which is the actual percentage of water in the bean. Both measurements are important for the safety, stability, and roasting of green coffee. Coffee that comes in with too high a level of water activity or moisture content may have quality and even toxin risks, and it is good practice not to accept these kinds of coffees.
Water activity is interesting to us when we roast coffee because it gives the roaster a road map on how to apply heat to remove the moisture content. Still with me? Both attributes of coffee can be measured using different machines. Water activity meters often use thin film capacitors to measure the water activity by measuring the air that surrounds the coffee itself in a closed container. The more water that is present in the air above the coffee in the container, the lower the energy capacitance of the bean itself, and the higher the water activity in the bean. Conversely, the less water is present, the higher the capacitance, and the lower the water activity.
Moisture meters have the ability to measure the moisture content of many different substances based on the principle that different substances have different dielectric constants—or the ability of a substance to hold energy. The same moisture meter can measure the moisture of wheat, soy, or coffee and it uses programmed dielectric constants based on the substance. (Side note: did you know that caffeinated and decaffeinated coffees have different dielectric constants and ideally have different programmed curves on a moisture meter?)
The most critical elements to control in green coffee mise en place are the temperature and humidity of your samples. Water activity will greatly differ based on the temperature and humidity of your samples due to the Ideal Gas Law—which in a simplified manner states that vapor pressure is directly related to temperature. Water activity is measuring the vapor pressure of coffee in a closed container, so if your green coffee is at 15°C you will get a different reading than if your green coffee is at 25°C. Very similarly, the moisture content will vary greatly based on temperature because dielectric constants vary with temperature.
What if your coffee actually had acceptable moisture content, but you had erroneous data because your sample was warmer than usual? You could be rejecting a lot or paying less for a coffee that was actually deserving of approval or a quality price premium. It isn’t fair to the supply chain before you if you don’t have consistent measurement practices! Out of respect and honesty, you are responsible for keeping the best practices that you can and sample preparation is part of that responsibility.
Everyone’s business and lab are different and I don’t claim to have the answer on the best way to prepare green coffee samples. What I do know is that to get reliable data you should create a procedure for preparing your samples so that you are always reading green coffee at the same temperature and humidity. A relatively simple change in your sample mise en place can have a measurable effect on your data accuracy, and, in turn, the entire coffee supply chain.