How age affects the ability to extract when it comes to coffee.
There are no shortage of myths surrounding how and why people brew coffee the way they do. One of the more elusive elements to a perfect extraction is the age (in days, post-roast) at which we brew the coffee. Armed with this fact, and all of myths surrounding it, we set out to get to the bottom of "age" and how it affects the extraction of coffee.
The general consensus is that coffee intended as espresso should be allowed to "rest" (read: age) for a period of days before brewing. This is in part to allow gases created during roasting to escape the bean which in turn allows for a more controlled and even extraction. After all, if gas is going out, there isn't much room for water to get in and extract the delicious solubles that make coffee the beverage we all love. This same principle explains the need for a pre-infusion, or "bloom" when brewing with manual methods. While none of this is breaking news or even particularly insightful, the questions it sparks and how we answer them can be.
Enter: The Question
How does the age of a coffee (post roast) directly affect our ability to extract that coffee, and how does this relationship change as the level of roast development changes?
To answer this, we designed a series of simple experiments and set to work collecting data.
Experiment #01, AKA: "The Really Old Aeropress"
For the first experiment we chose a single coffee, roasted it to a consistent profile, and took two samples from a specific roast every day for 30 days. At the time of the sampling, we sealed one in a one-way vacuum bag and performed this procedure with the other to measure the sample's potential extraction, or "solubility". At the end of the 30 day trial, we then performed the same procedure on all of the second samples, which had been stored in a cool dark place in their one-way sealed bags. The result was extraction data from the same coffee ranging from 0-30 days post-roast, as well as the data from each batch at 0 days (measured at the time the samples were taken). What we saw was the direct relationship between the age of a coffee and it's capacity for extraction within a fixed set of parameters. Eureka! Here's what we found.
The graph shows a positive relationship between the age of roasted coffee and the extraction levels within a fixed set of parameters. While the degree to which this effect was observed changed from sample to sample, the older coffee always* resulted in a higher extraction yield. *this is not definitive, and only based on a limited sample of a single coffee/level of roast development.
With this realization in mind we took a closer look at degassing itself (the likely cause of these results) and how it specifically affected extraction in relation to age.
Experiment #02, AKA: "Pulses"
Our second experiment focused on how age affected the solubility, and at what point throughout the extraction this effect was most pronounced. We took samples of several coffees from specific roasts to test both at 1 day of age and at a variety of extended aging periods up to 30 days. We then performed this procedure in order to measure the amount of extraction taking place at 6 different points throughout the extraction. This was similar in many ways to separating an espresso shot into the first, middle and final thirds in order to highlight the taste of under/over extraction, except with 6 samples each time. The results were then graphed as a curved line (%TDS vs PULSE #). Here is an example of a typical Pulse Brew of a medium roasted coffee 6 days post-roast:
The graph shows what we have come to expect; that is, an initial period of lower extraction (due to rapid degassing) followed by a peak extraction rate which then tapers off towards the end of the brew. This in itself confirms the validity of a pre-infusion, or "bloom" when brewing manual methods. Full extraction rates won't begin until the coffee has fully degassed, and without a pre-infusion to facilitate this rapid degassing, a significant portion of the brew water can flow through the coffee bed before extracting to it's full potential. The result is an unbalanced and chronically under-extracted cup.
As the age of the coffee we tested increased, this initial effect was reduced and the need for a dedicated pre-infusion became less important to the overall extraction yield and balance of the cup. Here is an example of the same coffee tested at 11 days post-roast:
This data reflects what we have become familiar with, specifically, that as the age of a coffee increases the initial degassing becomes less dramatic and shorter in duration.
As we extended the age of the tested coffees further, we noticed a continuation of the same flattening trend in the extraction curves. At a certain point in time, which varied from coffee to coffee, but almost always occurred*, the initial pulse would achieve a higher extraction yield than the 2nd pulse, leaving a consistent downward curve from beginning to the end of the brew. Here is a graph of that happening with the same coffee tested at 17 days post-roast:
This result was at once predictable, but also enlightening in it's implications. It showed us that after being roasted coffee continues to degas until reaching an equilibrium point at which the effect of age, specifically the extraction yields as a result of degassing, hit's a baseline and remains constant. This was further supported by testing coffees at 45, 60 and 90 days post roast.
Enter: The Answer (and a lot more questions)
Through a lot of time, even more data, and some suggestive results we've come to some conclusions about the effect age has on the solubility of roasted coffee.
- Pre-infusion plays a vital role in the balanced extraction of coffee less than ~21 days old.
- As coffee ages, the majority of the extraction happens earlier on in the brew cycle.
- Correct for the age of fresh coffee by:
- pouring more pulses
- decreasing the size of the first 2-3 pulses
- increasing the time between pouring them.
- Correct for the age of older coffee by:
- pouring fewer pulses
- increasing the size of each pulse
- decreasing the time between them
In addition to these conclusions, it has led to some further questions that we will endeavor to answer in the future through more testing and experimentation.
- Does this same phenomenon apply directly to other brew methods including immersion methods and espresso?
- Can the 'baseline date' be accurately predicted using green coffee density and agtron roast levels?
- Can the roast profile significantly impact the effect age has on the solubility of a given coffee?
- How does the growing region, altitude, varietal, processing, density or screen size affect the solubility of coffee as it ages?
We'd love to hear your feedback and response to our tests and especially want to hear if there are any conclusions drawn from this data that we haven't discussed here. This really is just the beginning of these sorts of experiments and we'd love to incorporate any and everyone into the process - after all, we're all on the same side of this one, that is, the side of better coffee!
Until next time,