So a coffee walks into a sauna

11 minute read

“Water water everywhere, nor any drop to drink”

Samuel Taylor Coleridge

A few years ago, I was at a winery in Napa, tasting through a vertical (a flight of wines from previous years—in this case, the past six years), relishing in the fact that I was experiencing a time capsule of the growing conditions and harvest and winemaker and cooperage and so on from each of those wines—and wishing we had something comparable in coffee. Wine lives in a bottle, capturing a moment in time, ready to be revisited and shared in the future. Sure, yeah: George is dipping his coffee in carbonite, or whatever (I know), and before mainstream media decided he said we’re making espresso wrong (he didn’t), Christopher Hendon investigated freezing coffee. But primarily, in coffee, we deal in ephemera. We control little, and understand even less.

What we taste exists only when we taste it.

From fermentation—which most often occurs spontaneously and is subject to native microflora and weather and hygiene and ripeness; to roasting—where we mostly measure the air surrounding the coffee and the metal housing it rather than recording a repeatable history of what the coffee itself experienced; to grinding—where a seed shatters without uniformity, and where the burrs wear and motor heats and seeds age as you go; to espresso—where our machines (other than, mercifully, the Decentnot an endorsement, just a fan) measure water temperature at the boiler and pressure at the pump instead of where the fucking coffee is (!)—we’ve built a structure around the sanctity of inconsistency, and, to the point of this post—ephemera.

You taste a coffee once—and then it’s gone. It can’t be repeated. 

It’s Snapchat, but for flavor.

It’s part of the romance of what we do, I suppose, and part of why so many coffee nerds are also into so-called “natural wine.” (Spare me your alcoholic salad-dressing or liquid-bread and pass me a well-made Chinon, thanks.) The idea that we participated in the creation of this thing that lived and was beautiful and then ceased to exist…it’s, well, romantic. (I prefer: narcissistic.) It’s the ostensible craft of being a barista, a skill best described as “overcoming the limitations of equipment and understanding to produce something delicious somewhat reliably.”

It’s part of the magic of coffee: Coffee is spontaneous. Chaotic. Alive.

And frustrating as hell.

When I was a young coffee buyer, I attended a cupping in Kenya that changed the way I understood this phenomenon—and understood coffee. The exporter prepared a cupping of 28 different coffees, hoping the purchasing group could collectively decide on a few lots to consolidate into a container for export. One coffee stood out: I scored it a 93—the highest score I’ve ever thrown down on an SCA score sheet (“Use the full scale, Feran” I hear you say). The coffee ended up going to another buying group, but I tasted it once it landed in the U.S. By then, naturally, it had fallen off—to the point that it barely qualified as specialty.

But that’s not how I remember that coffee. I remember it in its halcyon days. That coffee was fucking delicious. I’ll remember that coffee for as long as I can remember coffees.

It was there, but then: it was gone.

Specialty just means clean, sweet, uniform and 80+, y’all. I don’t know where so many of you got this 84+ shit. And that “sweetness” attribute score assumes that a coffee has it—you aren’t grading intensity. You literally only mark that box if sweetness isn’t in a cup, which pretty much only happens when you have other taints and faults like hard cup, non-coffee flavors (like kerosene) or extremely extremely past crop coffee.

It happens—all the time. A coffee jumps out from the cupping table—dimed out, vibrant, viscous, intense—while everything else is acoustic night at the dive bar by comparison. And then the exporter or producer tells you that it’s not done drying yet, that it’s still at 16% moisture—and inevitably by the time it hits the water it’s just not the same coffee anymore.

You just had to be there. 

There’s something about fresh, high-moisture coffee that is transcendent. I like stable coffee, and 90% of my buying is in pursuit of stable coffees so that I can buy more, bigger positions from fewer producers.

But high-moisture coffee isn’t stable. I wrote about that in my last post, but I left something critical unsaid: the role of lipid degradation in coffee quality. Overtime, as coffee sits, the fat content is affected by lipase activity and splits into free fatty acids (FFA), levels of which are inversely associated with cup quality:

As coffee sits, levels of FFA increase over time—faster when stored in conditions of higher relative humidity or moisture content.

Braz. J. Plant Physiol., 18(1):201-216, 2006, “The lipid fraction of the coffee bean.” Karl Speer and Isabelle Kölling-Speer.

In other words—as coffee sits, FFAs increase in concentration, phospholipids decline, antioxidants oxidize and decrease, and cup quality decreases. This relationship is accelerated by levels of high moisture, high humidity and high temperature: the same cocktail of ingredients that leads to increased microbial activity also causes an increase in enzymatic activity (of course).

So that’s why we dry to stability—even though high-moisture coffee pops when it’s fresh, like my dreamy Kenyan lot. We need coffee to hold up for long enough to have value once it’s been stored, milled, shipped, stored, and roasted, months and months later.

Is it possible to capture the magic of fresh coffee at our roasteries? Maybe without doing some profoundly stupid shit in the name of capitalism like air freighting coffee every week from Colombia to Brooklyn? 

What if we could do that—what if we could recreate that experience so that it didn’t have to just live and die on that coffee table, some 22-hour flight across the world?

What if we rehydrated coffee to “fresh coffee” levels? Would that even work?  Dry January is thankfully long since over, so: Hear me out.

I’m not the first to think about this. In fact, it’s an idea I first heard about from Tim Hill back when he was at Counter Culture—an access road in some circuitous conversation on the topic of fade and freshness and how coffee just off a drying table will pop on the cupping table. He said he’d played with rehydrating coffee a bit at CCC and that the effects were “interesting” but that it was difficult to execute operationally.

The idea stuck with me.

Sometimes I fixate on things (If you know me personally, you know that this is true). Moisture, and water, and the end times—it’s an easy feast for an anxious mind. Climate change is poisoning or drying the well depending where you are and I can see the 11th largest lake in the world from the front steps of my house. 

I give a damn about water: Water is life.

Low moisture is a problem for roasters; we associate it with coffee that tastes flat, less complex, and grain-like. Almost certainly, this has to do with both reduced vapor pressure in the seed during roasting as well as reduced rates of Maillard reactions that require availability of moisture. But some of that is relative—stable coffee at 10.5% might taste less vibrant than fresh, unstable coffee at 12.5%. What is “high moisture” in one context is delicious and desirable in another.

It’s kind of bizarre, when you stop to think about it: we pay to ship water to our roasteries, typically anywhere from 9-12.5%, by weight. It’s not unprecedented or unheard of to ship water—after all, there was once an industry dedicated to harvesting and shipping giant blocks of ice around the world (and what coffee buyer doesn’t appreciate Latin America’s love for ice cream, which is a result of the ice barons’s marketing campaign) and Nestlé continues to deplete fresh water in Florida to package it in shitty plastic bottles and send it around the world (or to places like Flint, Michigan, which is still without clean water).

In a standard shipping container, you might have 320 sacks of coffee filled with 60 kg of coffee each. That’s 19,200 kg in total, about 11% of which is moisture (this calculation is dictated by volume since 19,200 kg is well below ISO’s weight limit per TEU, which was updated in 2016 to 36,000 kg). In other words: 2,112 kg of that shipping container of coffee is actually water. And—since we’re talking specialty coffee here—that coffee is shipped in a barrier bag like Grainpro, keeping that level of humidity relatively stable and trapped within each bag.

Containerization—and especially the shipping container, which was invented in 1956—changed intercontinental commerce and, frankly, made the explosive growth of coffee consumption possible by making it more affordable than ever to import goods from other countries. Gone are the days of shipping coffee from Sumatra in barrels, exposing it to the conditions of wind and weather—well, unless you’re into that sort of thing.

But what if we could dry coffee to a stable point—say 10%—and then add that moisture back right before roasting? What if we could Lazarus that coffee and make it come to life again and taste the way we remembered it tasting back in Suaza?

In October 2019, Roast Magazine published an article about the connection between seed viability and quality, something that has been studied extensively in other seeds and which has some discussion in coffee. In coffee, we’re not as concerned with the rate of germination but rather the theoretical viability of the endosperm (rather than the embryo). This is why we dry coffee to 10-11%, as we know that drying to that level will still retain viability of the endosperm (NB: there has been some investigation into the viability of seeds dried below 10%, but generally the dehydration was done rapidly with high temperatures—130°C!!!—even though we know anything over 40° will likely kill the seed. These studies also tend to use a classical focus on the embryo and germination as their focus rather than looking at the endosperm specifically so it’s difficult for me to take their word for it. Seriously, who the fuck designs these studies? Breville and UC Davis?).

(Another note, other studies sort of look in the right direction but somehow talk about a loss of glucose and fructose without mentioning microbial activity as being implicated in fade—they just accept the loss of sugars as a fact and skip past causes, instead focusing on “storage conditions” and moisture content. Again: I fixate on things)

But: I can’t seem to locate a study exploring what happens if you dehydrate coffee to stability and then rehydrate it again.

My business partner at Phoenix is a talented ceramicist and observes his pottery wheel with the reverence and fervor of a newly saved soul at the altar, going so far as to build a pottery studio in his basement (more often than not meetings are set around his kiln times—hard to blame him). He told me about the technique of “reclaiming”, where dried scraps of clay are rehydrated so that they can be used.

And in bread baking, of course, where the oven follows a similar pattern to roasting, beginning with a tremendous amount of energy that slowly declines over the course of the bake—steam is integral to proper development of the crumb and crust and is, in fact, injected into the oven at the start of the bake and then mostly evacuated by a third of the way through.

So what would happen if we rehydrate coffee, injecting some steam in the oven—so to speak?

I took coffee that landed a month ago from a group of 15 producers I buy from in Huila, Colombia and analyzed its physical characteristics. At 10.2% moisture and 0.49 water activity and with 0% fully fluorescent seeds, only 1.7% UV speckling, and a bulk fill density of 0.72 g/mL, this is a highly stable coffee. I’d purchased a large enough position to last for a full calendar year as a result, as it is a core component of Phoenix’s blend for cold brew. In two separate cuppings, I’d scored the pre-ship an 85 and the arrival an 84.75.

I took a sample from the arrival lot and rehydrated it to two other moisture levels — 12.5% (the upper limit of ICO’s export standards) and 16.8% (close to one of the inflection points in most drying systems at wet mills and a place where a type sample of a coffee is often first tasted).

To accomplish this, I placed the first sample in a humidifying chamber at a constant temperature of 30°C with an equilibrium relative humidity at 70% and periodically checked the moisture content. 

Source: Coffee: Growing, Processing, Sustainable Production

You can see from this chart (borrowed from Coffee: Growing, Processing, Sustainable Production), this puts me on the absorption curve targeting a final moisture of 12.5% for green coffee stored in this condition. Within 48 hours, I had reached that target humidity, removed the coffee from the humidity chamber and placed it in a closed container to stabilize.

Sidenote: this is also why it’s difficult to dry coffees to stability in parabolic driers or marquee driers that don’t have the ability to open their sides—the relative humidity and heat both stay at high levels so the coffee ends up around 12-12.5% moisture, even if it (initially) reads lower on a moisture meter (remember that readings on a moisture meter are an average. Tracking weight of a fixed volume of coffee is a much more accurate field measurement). This encourages the growth and metabolism of microorganisms as well as fat oxidation and leads to early fade, as discussed above and in the previous post. Proper drying of parchment coffee requires sufficient airflow. Otherwise, it’s a coffee sauna.

I then took an additional sample and quickly submerged it in cool, filtered water; allowed it to drip dry; gently patted it dry; and placed it in a closed container. Within a day the moisture level had stabilized to 16.8%.

The reference sample and rehydrated sample at 12.5% look quite similar (difficult to tell from the above) but the rehydrated sample showed more uniform color. The Rehydrated sample at 16.8% had a deep green-blue color that pretty much looked like wet hulled coffee, but without the squished and broken seeds.

When I achieved the target moisture levels, I pulled and isolated 50g from each of the prepared samples as well as the reference lot at 10.3% moisture for roasting on an IKAWA sample roaster.


I cupped the three samples blind using the standard SCA method a day after roasting. Here are my notes, with the codes replaced by the identified coffees for ease of reference:

Dry: Intense floral bright apple stone fruit. Similar to 12.5% but more intense.
Wet: Grape and corn and fresh
Cup: Bright, primarily citric acid profile. Long finish but somewhat drying and vegetal. Notes of caramel, apple, cherry, sweet, slightly dry, a little light body but good structure.

12.5% moisture
Dry: Orange cherry cocoa almond nut. Nuttiest of the three.
Wet: Grape and grainy
Cup: More intense than reference,  acid is subtly more tangy. Cherry dominates. Slightly higher body. Caramel, sweet.

16.8% moisture
Dry: Sweetest fragrance, most vibrant, stone fruit, winey, grape, cherry
Wet: Fruit forward, sweet
Cup: Sweet, caramel, cherry, guava, grape, juicy, body improves as it cools, tangy acidity, most complex

So I have to agree with Tim, here—it’s interesting. Operationally, submerging coffee in water and allowing it to drip dry and homogenize in a barrel for a day doesn’t seem any worse or more tedious than the process of roasting from frozen (and uses signficantly less energy than maintaining a giant walk-in freezer)—but is it worth it?

You know this already, but: I like the idea of drying every coffee to 10% moisture. I do.

When you dry to 10% moisture versus, say, 11%, there’s more coffee seeds per bag—in a 60kg sack you get an extra 600g of green coffee versus 11%, which, in a container is an extra 192kg of coffee. That’s the equivalent of more than 3 bags of green! Plus, assuming you don’t rehydrate it—yields are higher out of the roaster because of less roast loss—meaning you need less labor to move, lift and roast coffee. You’ll have lower in-bound freight costs since you don’t need to pull coffee from the coast as often as a result (or wherever your green warehouse is because for some reason we have landlocked green warehouses in the Midwest); and you’re not paying financing on what is literally just excess water in the coffee you have stored there. And the shelf life is longer.

But at the moisture level of fresh coffee—it just tastes better.

Why can’t we have both?

I remember the way that coffee tasted in Kenya—hail Satan—and I want to taste that again. What could we have done differently? How could I taste that again?

Let’s dry to 10% and get that coffee to stability with all of its associated benefits and store it in GrainPro and control the humidity and heat in our warehouses to extend its shelf life—but let’s not freeze it. Maybe because it’s winter here at the 41st parallel but I’ve had enough of the cold. On grey February days like this I want to live in the sauna. And maybe my coffee does, too.

And maybe that way—even if we don’t freeze it, and even if we can’t keep it like a time capsule in a 750mL bottle in the cellar—we can recreate that first moment when we tasted that coffee and share it with the world.

We can make it more like the first time, every time.


That's just, like, your opinion, man

  1. This is very interesting, but I got caught up on one little detail; did you adjust your Ikawa roast profile for the different moistures or do you believe the machine sufficiently compensates?

    We have an Ikawa and varying moisture has often reduced my confidence in the fairness of comparisons.

    1. I wanted to eliminate roast profile as a variable—while acknowledging that rate of reactions will vary based on moisture content (but that’s always true since they’re linked, and that’s also sort of the point). The profile I used is pretty forgiving and robust—I use it across all of the coffees I sample over 0.70 g/mL density, irrespective of moisture. Happy to send you a link if you want to try it out 👌

  2. A compelling read, and intriguing report. Does a moister green coffee transfer heat more efficiently to the core of the seed? The better development would be a boon. Or is that wishful thinking?

    Your rehydrating method has me curious if a particularly difficult Kenya I have, which seems to go from underdeveloped to roasty in about 10s (and which window I can’t consistently locate of control) might become easier. Again, probably wishful. But I’m a home roaster, so nothing much at stake.

    Thanks for posting this!

    1. That’s pretty much my theory—faster thermal transfer and faster rate of reactions. Also worth noting that higher humidity air holds more energy – part of why we need to approach crashy Kenyans differently, imho (maybe I’ll have to write about that too. How endothermic flash impacts a roasting system).

      It’s an easy test and you can basically approach your roast the same way – give it a go!

      1. Hi Christopher,

        I gave this a shot and cupped the two reference lots all of 5 hours out of the roaster. Seemed to us that there was a fairly noticeable difference; for our coffee, the reference seemed to have more body, the rehydrated coffee was sort of cleaner, more transparent and had a higher quality of acidity.

        I tried to use the same profile for both and it looked like the rehydrated coffee lost about as much more weight as the water that it had absorbed.

        I’m wondering if part of the effect of this is to delay the onset of browning reactions, since the additional water needs to be driven off before they can happen. What surprises me, though, is the effect on acidity.

        Anyway, I’m not really much of a roaster, so I might repeat the experiment and be a little more methodical in doing some moisture maths. For the moment, the samples are sitting vac sealed, waiting to age a bit to hit what should be their optimum consumption window.


        1. Thanks for sharing your experience! Curious how the vac sealed samples perform — they should be even more homogenous, which may help a bit.

          I suspect you’re right about delaying the onset of browning reactions, though once they began I believe they probably occur more rapidly (and thus, development occurs more rapidly). Important to consider that not all types of acidity degrade during the roast — acetic acid (which is also produced during fermentation) increases over the course of the roast as small to medium chained polysaccharides (including sucrose) break down (acetic acid levels can increase up to 25x over the course of the roast) which greatly increases perceived acidity and fruit character. Lactic acid, too, is developed during the roast (with sucrose as a precursor for lactic acid, too).

  3. I loved this read, thank you for sharing this insight.

    I too have had similar experiences with coffees at origin that were forever beautiful in my mind only to be a ghost of their fresh selves when you cup any other sample after it’s traveled. I always attributed it to fresh v dried fruit, they just don’t taste the same but are basically the same thing.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top