Machine-Washing, Espresso-Shooting


“Machine-washed” is a confusing term if you ask me, commonly used to describe one of the more dominant processing techniques in Costa Rica. It’s partly confusing because “washed” normally implies fermentation, a step used to break down the fruit/mucilage layer making it easy to remove from the seed. No, there isn’t mechanical fermentation with “mechanical washing,” rather the fermentation step in the wet-processing technique is completely foregone, using a machine to remove the cherry and most of the mucilage in one fell swoop, producing parchment coffee with only a thin layer of mucilage intact when drying on patios, in greenhouses, or in mechanical drum-type dryers (guardiolas).

So, isn’t this the same as “pulp natural” processing used in many other coffee-producing countries? Sort of. Pulp natural coffee can be produced with both de-pulpers and de-mucilage machines, the common element being that only the cherry is removed, leaving behind the entire mucilage layer during the drying stage. But with machine-washing, there is control over how much of the mucilage is left intact, which can greatly affect the influence processing technique has on the cup: The less you leave on, the greater the transparency. In Costa Rica they’ve given names to correlate wiht the amount of mucilage left on the seed roughly correlating to percentages; these are the infamous “honey” coffees: black and red honey=50% or more mucilage; yellow honey=20% – 50%, and white honey/mechanically washed less than 20%. Measurements are of course imprecise, but they can come pretty darn close.

So how does this relate to espresso? Well, like pulp-naturals, the pulp left on the seed tends to produce heavier body, restrained acidity, and slightly-fruited sweetness, positive attributes for espresso shots. For the machine-washed/white honey category of “20% or less” there’s quite a bit of room for variation, and so no surprise that profiles of coffees categorically produced the same way (i.e. white honey/mechanically washed) can show vastly different cup attributes. Such is the case for the two current Costa Rican coffees on Coffee Shrub, both mechanically-washed and from two different micro-mills; Chirripo Finca Jose and Helsar Magdalena Vega.

Finca Jose is from a brand-new Chirripo mill, the effort of two brothers who inherited a few large coffee farms, that between them, produce enough coffee where a “micro-mill” operation makes sense rather than pay to have their coffee processed by someone else. Magdalena Vega’s coffee on the other hand, is from the well-known Helsar mill in the West Valley, one of the original micro-mills, known for producing astonishingly clean mechanical-washed coffees. One look at these coffees side-by-side, and the variation is quite obvious. Jose’s coffee appears to be close to the “20%” mucilage threshold, the raw coffee dawning red to brown hues. Vega is a fairly uniform green coloring, closely resembling actual “washed” coffees. The cup profiles echo these differences too, Finca Jose a fruit-forward, juicy bodied coffee, and fairly complex in comparison to Vega, which is characterized by balance, “classic” bittersweet notes, and articulated sweetness. No surprise they offer very different espresso experiences too.

Both coffees are made up mostly of caturra cultivar, and grown at very high altitudes (1700 – 2000 meters). This makes for fairly dense coffee beans, and they need to be handled in the roaster as such. Both of my espresso roasts are Full City, that is, roasted to a level of darkness where shade-uniformity is achieved, but not quite encroaching on 2ndcrack territory. Like pulp-naturals grown at similar altitudes, these coffees can take the heat up front in order to hit the desired time markers, which for my sample roasting is 3 to 4 minute yellowing, 7 minute first crack, and pulling around the 10 minute mark for FC. I have to up my PID controller a few degrees after warm up to make sure I don’t lag on these numbers. When roasting in our Probat L12, we’ll bump up our initial charge setting about 20% (from 1wc to 1.2wc on our controller), coercing a touch more heat on up to 390, and then reducing back to just below the initial charge setting (not cutting heat altogether) as we roll into 1st snaps. This last part keeps the roast from stalling, and at a rate of rise that will progress the roast all the way to the desired FC temp, for us around 428 (It should be noted that we keep full air-flow through the drum for the first 5 minutes of every roast in order to cool and roast at the same time, after which we cut air completely until the batch is pulled).

OK, so we’ve roasted both the FC, let the roasts rest a few days (96 hours for this assessment), and are ready to run them through our espresso machine. First I’ll say that it took me a couple tries to get the Magdalena Vega dialed in. I don’t think this has much to do with the coffee itself, but more the inner-workings of an ultra-sensitive K3 we have in our lab. I of course made sure to try every shot that passed through our machine – even the terrible ones. I have to say, my initial 15 second 4 oz-er as well as the 45 second 1 oz-er showed promise. Once I dialed it in, the two “good” shots pulled were velvety thick, and produced a visually-appealing level of crema. After cooling for a couple minutes and a quick stir, I take my first sip. Vega will appeal to those who lean toward “classic” profile espresso – that is, flavor profiles revolving around a tart and sweet sensations up front, and chocolate richness from middle to finish. I will even go as far as to call the espresso “mild,” in that it doesn’t taste like an intensified version of all the characteristics found in the cup. The intensity is there, but a much more restrained version than most.

As you might expect from the Finca Jose cup review (and what we’ve laid out in this article), fruit-forward characteristics reverberate in the shot. There is a sustained chocolate note too, that is mouth-coating, nagging at the palate long into the finish. It’s heavy, bitter and sweet, and provides a grounding quality for top notes to play off. Fruited flavors go from cherry to berry, and brightness is also in the realm of tartaric/grape-like. Though roast times were nearly identical, this coffee was roasted just a tad lighter than the Vega, I’m guessing partly due to higher altitude (his farm is just over 2000 meters), as well as the fact that there appears to have been more fruit left intact post de-mucilage stage.

Both coffees will stand up to 2nd snaps no problem. Tapering the heat just before 1st crack is crucial to avoid rushing through this important leg of the roast, drawing out sugar caramelization as you also invariably introduce heavier roast tones into the cup profile. For Vega you can expect development of chocolate roast tones, bittering cacao slightly overtaking sweetness, but not altogether. This type of espresso profile will work great for espresso drinks. And this extra phase of development will round off some of the fruited “edge” of Finca Jose’s coffee, also bolstering the ever-present cocoa roast tones. With this one, you can expect a complex espresso shot no matter what.

What’s interesting to me about running these two side by side, is that they help illustrate the fairly wide allowance within the “machine-washed” processing category. Just by looking at the green coffees it’s apparent that they are different – even though both Caturra processed in the same manner – and these visual cues help inform roasting decisions as well as give a general idea of what to expect in the cup (i.e. cup clarity, fruit, etc). I think this is especially true for the mechanical washing method, where the “cleanest” examples are near-completely stripped of fruit, those closer to pulp-natural at the other end of the spectrum. Vega and Jose exemplify this wide range, both representing the top-tier of pulp-natural style espressos. We have a few bags left of each, Chirripo Finca Jose and Helsar Magdalena Vega, on the Coffee Shrub Green Coffee List.