Sumatra, whose short supply was a thorn in the coffee-drinking worlds’ collective side for the past year or so, is finally rolling its new crop into port. This summer arrival from the highlands south of Takengon and Danau Laut Tawar caught our attention with its juicy mild fruit character and unique herbal undertones, elegant accents to an otherwise clean and classic Sumatran flavor profile.
The coffee comes to us from a smallholder group near Atu Lintang, which is then collected, milled, and exported by the Jagong dry mill. Jagong is operated by Irham Junus and his children Ina and Andi, who have maintained a close relationship with Royal over the years. Their strict adherence to ripe cherry collection is one excellent example of their attention to detail and rigor in quality selection.
The greater region of Aceh, Sumatra’s northernmost province, is the source of “Mandheling” coffee and is sometimes referred to as Gayo or Gayoland in reference to the local Gayonese ethnic majority. Coffees from this region, like this Crown Jewel, are frequently cultivated by refugees of regional conflicts and earthquake survivors, clinging to a traditional livelihood that involves preparing coffee using Sumatra’s all-too-famous wet hulling process, sometimes called semi-washed (not to be confused with semi-washed coffees from the Americas). Known locally as “Giling Basah”, the procedure usually entails some form of pulping, followed by a brief drying period to reduce moisture to anywhere between 20-45%. The coffee is then delivered to the mill, where the parchment is removed while still damp, and the coffee completes its drying as the raw green seed.
The practice arose from Indonesia’s earliest days of cultivation under the Dutch colonizers, and is as pragmatic an exchange as exists in the coffee growing world, one that minimizes labor and time at the front end of the supply chain – an important consideration when labor and time aren’t being compensated. This method leaves its mark on the distinctive jade-like color of Sumatran coffees, as well as their funky, earthy flavors, unmatched elsewhere on the globe.
Having come from a fairly rigid and fully-washed-focused school of thought, grading Sumatran coffee has come with a bit of a learning curve and adjustment period. Indonesian coffees arriving at above 12.5% and water activities in the mid and upper 60s is commonplace, a direct result of wet-hulling and a damp harvest season. These coffees also tend to be of lower density as a result. One surprising characteristic of this lot its its screen size – usually quite large beans, this Atu Lintang is spread widely across the top four or five screens, making it closer in distribution to a Central American SHB EP than a traditional Indonesian offering.
Three prototypical coffee varieties are present in this Sumatran coffee, each with roots in the Pacific archipelago. Indonesia (by way of Yemen and Ethiopia) is the source of the vast supply of the world’s cultivated Arabica. A single plant, pilfered from Yemen’s port of Mocha, eventually gave rise to Javanese Typica, which in turn populated the Western Hemisphere. It was the source material (by way of the French, who received a plant from Java as a gift from the Dutch) for Reunion Island’s Bourbon mutation as well. At the end of the 19th century, much of the Pacific’s Arabica was obliterated by rust, and Robusta was introduced to replace the dwindling Arabica supply. On the island of Timor (now split between Indonesia and the independent nation of Timor-Leste) against all genetic odds, Robusta crossed spontaneously with Arabica and produced the parent hybrid to the subsequent family of disease-resistant Arabica selections, Catimor. Catimor, in its many forms, can now be found throughout the coffee-growing world.
With Jen out of the office for a few days at the Roasters Guild Retreat, I took up the reins on the Ikawa this week, and promptly rearranged all the furniture, figuratively speaking. Kicking up the fan speed to the new “open” setting updated with the firmware v.23 setting, I resolved to reinvent some roast profiles.
I took an archived profile Jen worked on, much longer than we’ve been roasting lately, at around seven minutes total. This roast (blue) produced a classic but unremarkable Sumatran profile, earthy and chocolatey with some mild roast notes and not much unique character to speak of other than a hint of grape soda
Noticing the rapid RoR at the beginning of the previous roast, and making one or two assumptions about the need to extend the drying stage at lower temperatures for high-moisture wet hulled coffees like this Sumatra, I decided to try something dramatic. I forced the turnaround temperature below 200F and used a down-trending fan-speed to compensate for the loss of moisture weight throughout the roast. A significantly different style of roast than the first, this second attempt (red) showcased a juicier and more unique flavor profile with mango, black tea, citrus, and mint all playing supporting roles.
It’s a malleable coffee, one that might take some finesse to perfect in the roaster, but a rewarding cup nevertheless.
Roast 1: 6:45 420 af\
Roast 2: 7:00 CKornman Sumatra
My favorite Indonesian coffees have two dominant flavors in the cup: a deep sweet chocolate and a juicy citrus quality all the while having a syrupy mouthfeel. This coffee fits the bill. I started with a medium high charge temperature of 382F and applied heat just after the rate of change peaked on the screen at 1:55. Yellowing began at 3:10 which was sooner than I thought so I turned the heat down slightly to extend the Maillard stage and the total roast time. At 6:14 the roast was at 381.5F and I could hear a few snaps in the drum. I could also see that the rate of change was beginning to drop too quickly and so I turned the heat back up to 3 gas.
First crack came early in the roast at 392F. There was no major “dip’ in temperature so I turned the heat down 43 seconds after first Crack began and once more 1:10 after first crack to 2 gas which is my minimum setting. My final end temperature was 409F and there was 1:27 minutes of post crack development time.
On the cupping table I had a set goal to round out the acidity of this coffee so that the sweetness of the fruit and sugar browning notes would shine through. This roast worked perfectly and I would not change a thing if I was to roast it again. There was a vibrant acidity that was uniquely Sumatran in character of grapefruit and cedar, all complemented by a large heavy body that was very sweet like a tootsie roll.
For this roast I used two probes: the TJ36-ICSS-18U-6-SB from Omega for ET (Environmental Temperature) and an a12031600ux0143 from Uxcell for bean temperature (BT). I also broke out the Yocto-Thermocouple, a small device that relays the information coming from the probes to a computer of your choice. In this case, I used a Raspberry Pi 3 Model B running Raspbian Stretch. There are some excellent guides to set this system up, which I will detail in a later article!
The ET probe fits perfectly into the top left attachment point of the green bean chute. I removed the nut, and placed the probe most of the way in, but not completely flush. The BT probe screws right into the face of the roaster, directly below the trier.
While not entirely plug-and-play, I was able to get the system working. I placed the two leads in the appropriate ports on the Yocto-Thermocouple chip, opened up VirtualHub at the default 127.0.0.1:4444 address in my web browser, and was able to see that both of my probes were giving readings by opening up the API. If you want to see a visual representation of these readings, you can get Yocto-Visualization, a free software that works with this setup. One minor point: you’ll have to stop VirtualHub with the command ‘sudo systemctl stop yvirtualhub.service’ if you want to use Yocto-Visualization. Only one of these programs can run at a time.
For roasting this elegant Sumatran coffee, the thermocouples came in very handy. I started my roast off with the standard settings: 9A power, back open to stymie airflow, and a charge temperature (as read by the ET probe) of 365F.
I knew that this coffee with its higher moisture content, would need a firm push. I didn’t close the back of the roaster until 3:00/230F, and didn’t engage the fan to 3 until 5:30/293F. At this point, the coffee began picking up a bit more speed, but I didn’t want this coffee to rush through Maillard. In service of that, I dropped the amperage to 7.5A at 6:45/320F. This had more of an effect than I was expecting, and I waited until 10:35/374F to increase the fan speed to 6. This coffee cracked a significantly higher temperature than the others at 11:45/388F, and I dropped the coffee after 1:15 development at 13:00/394F – I thought this roast was a bit long in the tooth.
It didn’t taste that way on the table or brewed, however. As you can see from Sandra’s notes below, this coffee garnered lots of ethereal and fruity notes. Cedar, lime, apricot, and raisin all made headlines, and the cup became significantly more complex on cooling. I am currently reaffirming my love for Sumatran coffee with this Crown Jewel.
Lately, I’ve been struggling to get coffees to taste as good as I know they can, wrangling grind sizes, pulse pours, and TDS readings to try and achieve some lofty idea of the perfect cup. Not so with this Sumatra. Rushed one morning, I decided to get a quick brew in before heading out of the office for an appointment. I used my tried and true V60 recipe; Jen’s roast had cupped beautifully and I was sure that the clean extraction of a V60 infusion would shine a light on exactly what this coffee had to offer.
With the EK 43 set to 8, I ground 25g of coffee and used 400g of brew water for a 1:16. After stirring the bloom and letting it degass for 30 seconds, I gently added water 100g at a time every 30 seconds or so with the Fellow Stagg Mini. The resulting cup provided an explosion of flavors, all complementary and well balanced. Notes of orange, hibiscus, red grape, and rose were met with cacao, cedar, almond, and honeysuckle lingered delightfully on the tongue.
Since the Jagong Mill coffee didn’t really need any improvement from the previous brew recipe, I chose instead to switch up brew devices. Pulling out the Kalita, I implemented the same recipe I had used for the V60 brew – 1:16 ratio with 100g pulse pours. This brew tasted like apricot, vanilla, and honey, with less of the tropical acidity present in the V60 but still had plenty of sweetness. This is a surprisingly forgiving and truly delicious coffee.