Coffee Fermentation: Shrub Tasting at 2017 SCA EXPO

Wet fermentation - this coffee is just about ready for a final rinseWet fermentation - this coffee is just about ready for a final rinse

This post-SCA Expo blog entry is a little overdue. Returning from SCA is usually a bit of a rush, coming back to a week's worth of offer and arrival samples to roast and taste, and a steady stream of producers and exporters stopping through before they return to their homes abroad. Needless to say, our intention of following up our Coffee Shrub SCA tasting and discussion with a "quick blog post" was sidelined just a bit. But you know what they say, better late than never.

Coffee Shrub hosted our first ever SCA event at April's Expo, a gathering that drew a surprising number of folks given the slightly cagey instructions to an offsite location. It felt a little like a secret warehouse party in this respect, and so a big thank you to all those who heeded our instructions, waited for a location confirmation, and ultimately walked the several blocks from the Convention Center to join us!

For those of you who couldn't make it, we focused our tasting around different processing methods, and discussed the role fermentation plays on taste, and ultimately price. We were lucky enough to have colleagues from different coffee growing regions to take part in the discussion, including Fredy Morales who manages Finca Rosma in Guatemala, Ture Waji from Mormora plantation in Ethiopia, and both Leonardo Henao and Pedro Eshavarria from Pergamino Coffees in Colombia.

Fermenting coffee without water in Timaná de Huila, ColombiaFermenting coffee without water in Timaná de Huila, Colombia

Each has a different first-hand experience of how processing affects coffee in their regions, and what became clear as we moved through the discussion is that there is no "one size fits all" coffee processing solution that magically adds value to coffee. There are a lot of factors for farmers to consider, especially the small holder, and what might be considered a contradiction to processing "best practices" in one area may be the most lucrative option in another.

One such "contradiction" is the practice of mixing days of harvest during the fermentation process. This isn't talked about a lot, at least not in a positive light, as the current opinion by most agronomists is that harvesting and processing each day of harvest separately is the best way to operate. This means picking and de-pulping the coffee same day, fermenting overnight to remove the sticky layer of mucilage, and then laying the wet parchment to dry on patios or raised beds.

For small farmers, this also means managing several tiny batches of coffee, which is a bit of a strain at the fermentation tank, and a lot to keep track of on the drying patio. Most folks only have a single tank, and so in order to maximize the output during this stage in processing, you can either wait to pulp cherry until there is enough to fill the tank, or as we found in northern Antioquia in particular, pulp as you go, mixing multiple days of harvested coffee in the tank to ferment before drying.

Typical depulper set up in Urrao, Antioquia - this one has two small tanks for keeping mixing of days to a minimumTypical depulper set up in Urrao, Antioquia - this one has two small tanks for keeping mixing of days to a minimum

One of our guests, Leonardo Henao, has devoted quite a bit of time and research into learning how the practice of mixing lots intersects with cup quality. Having a farm in Urrao, Antioquia himself, he's witnessed mixing by his neighbors, some of whom produce coffees of exceptional quality (one of them won COE in 2014). So he and his colleague Doriett, a phD student from the Universidad Nacional de Medelliín set out to better understand what's happening during the fermentation process when mixing multiple days' lots, and more specifically, find out what benefit there is, if any, to the coffee's cup profile.

During fermentation, bacterias break down pectins in the sticky mucilage/fruit layer that is stuck to the bean, eventually allowing it to slip away from the seed. This also frees up sugar molecules, which are then broken down by yeasts, and produce organic acids and oils as a byproduct. In this way, fermentation isn't being used to affect cup flavor, but as a means by which to necessarily "wash" away this layer of fruit before drying. Under normal circumstances, the fermentation is complete once this is achieved, and simply leaving a coffee for longer puts a potentially clean cup profile in jeopardy of turning sour. What Leo and Doriett found by watching Leo's neighbors process coffee, and through laboratory research, is that certain steps taken along the way actually slow the fermentation process, allowing the coffee to sit safely for a longer period in the tank, and often improving the resulting cup profile.

First off, it might help to give a quick explanation of what exactly we mean by "mixing days". In Urrao for example, farmers mix up to 5 days of harvested cherry before the final washing off of mucilage and drying. First they de-pulp the harvest from day 1 into their fermentation tank and add cold water, allowing it to ferment overnight. The next day they de-pulp harvested cherry from day 2 on top of day 1, changing out the old water with fresh, cold water again. This continues until their fermentation tank is full (usually 2 - 5 days depending on the farm size), allowing the entire lot to ferment one final night with the freshly de-pulped coffee before rinsing the mucilage and sending to the drying beds. On it's face, we'd expect this process to result in flavors associated with over-fermentation, fruit notes on the alcohol side, and perhaps water activity on the high side. But what we found is quite the opposite, and regional conditions and local practice are part of the explanation why.

Urrao is extremely high in altitude, 1850 meters above sea level at the valley floor, and cold, historically too cold to produce anything more than blender type coffees. A warming climate has actually benefitted coffee growers in this area, with annual temperatures right in the sweet spot for stimulating high productivity for the coffee trees (averaging 19.5 - 21 C). Yeasts and bacterias need to reach a certain temperature to react with sugars, and the cold climate and water slow this process down. Changing the water also helps slow things down, as each day when cold water is added, the PH level is in turn raised, reducing the acidity needed to break down the sugars in the mucilage. And finally, each day new coffee is added to the batch, new sugars are introduced, maintaining a healthy environment for the surrounding yeast and bacteria.

Measuring and comparing org. acid levels in different process batchesMeasuring and comparing org. acid levels in different process batches

For Leo's experiment they tested several process batches that differeed in number of days mixed, and overall time in the fermentation tank. The goal was to then measure the organic acids present in the final milled coffee, and of course, record cup scores and notes generated by a sensory panel of cuppers. The slide above from their final presentation illustrates the change in organic acid content, the low end being coffee processed without any fermentation (mechanical demucilage), and the long end being 5 days of mixing. It's clear that of the 6 organic acids measured, the greatest overall volume is at 3 days, all but lactic and quinic seeing a steady increase in volume over the first few days before dropping off precipitously by day 5.

So how does this play out in the cup? As you can see from the slide below, the overall score range is pretty tight: 80.5 at the bottom, and 86.38 at the top (4A). The two top scoring coffees on this graph are a mix of 2 harvest days. The lowest scoring coffee was not fermented at all, the mechanical demucilage method being used to remove the mucilage and then immediately taken to the patio for drying.

Cup testing the process batches - highest score is with 2 days mixing, and lowest no fermentation at allCup testing the process batches - highest score is with 2 days mixing, and lowest no fermentation at all

What this graph seems to say is that the mixing method is superior to processing each day separately, which when we're talking about the difference of 80 and 86 in this particular case, means a hefty financial gain to the farmer. But it's important to keep in mind that given different variables - climate, harvest volume, altitude, bean desnisty, water temperature, etc - the effect fermentation has on a coffee will also vary. In a growing region like Costa Rica for example, where milling equipment operates at maximum capacity daily, and mechanical demucilagers are the norm, mixing days makes no sense at all. The same can be said for other parts of Colombia, and even within Urrao itself, and so for the farmer it's important to consider cost of production and quality implications when weighing one processing method over another.

Leo's presentation gave us a lot to think about in regards to the role fermentation has on a coffee's quality, with particular attention to the unique situation presented by small holders in Urrao. It's a reminder that for coffee, as with most agricultural products, environmental factors affecting quality are often out of the farmer's control, and differ from one region to the next. Processing, however, puts some control back in the farmer's hands, still limited by the outside factors mentioned above. 

During this discussion we compared the coffees of Urrao to other Colombian coffees subjected to much shorter fermentation times. Unsurprisingly, some folks favored one over the other, and thankfully not everyone agreed. There really is no accounting for preference, and as coffee buyers it's important to remember that no matter how much we love a clean washed coffee, or wild and rustic dry processed cup, "best practices" are limited by region, environmental factors, and the economic resources available to the farmer (and more). And so for most tiny farmers it simply does not make financial sense to chase current processing trends in the hopes of selling their coffee for a higher price. 

This can be discussed in much broader terms as well - e.g. not just decisions farmers face with regards to the intersection of processing style and cup quality, but also cultivars to plant, conventional vs. organic farm practices, storage methods, and so much more. In each of these cases, there is no blanket solution to controlling coffee quality.

If you'd like a copy of the presentation slides, email me at FYI - there isn't a lot of accompanying text with the slides, and they will be most useful to those who were in attendence. 

-Dan Wood