Prufrocker: The Blog

Brewing’s Third Species

Don't buy a machine. Be a machine!

by  |   |  Coffee Science, Equipment


When teaching a Brew Methods lesson, we usually set up Immersion and Drip as the two species of brewing and then spend half the class on one, half on the other. Last weekend however, we only dedicated two thirds of the class to these two. Because now there’s a third species of brewing! The lesson plan was to lead towards a topical finale to the course where we would make sieved coffee and coffee shots. This lesson was peppered with references to the espresso machine as ‘super brewer’; the ideal. We must try in vain to emulate its many virtues. Any point where we were slipping below the brewing precision of our espresso machine, we would flag these moments and try to improve our consistency and stability. Home brewers were encouraged less to go out and buy a machine and more to be machines!

These were the topics we would cover.

  • Water wastage
  • Thermal mass
  • Liquid retained ratios
  • Flow restriction
  • Speed
  • Multiple orders
  • Appearance, turbidity
  • Aroma (intensity)
  • Texture/Filter material
  • Craft. See Varney

When Bek Freeman headed up our brew-bar, she instituted an inclusive ethos that we wouldn’t use methods that couldn’t easily be reproduced at home by our customers. Bek had been an every day regular at the Penny University and the kodawari of Tim Williams’ hand brewing had a big influence on her. This placed a great deal of emphasis on the raw materials (Bek was our first barista to really zone in on water quality), and certainly made Brew Methods a hands-on and easy class for us to teach.

Then Colin Harmon started using his Uber Boiler to make Chemexes straight from the font; something that seemed very slap-dash to us at the time and we had to accept, this was kind of the Uber’s raison d’etre. The +-0.1deg of temp stability with a mere 1deg offset from the boiler (John Gordon has tested this at Prufrock) has been (no pun intended) thoroughly capitalised on since early in the Third Floor nightclub days. We were, and are still a bit committed to the pouring kettle but to get more thermal mass, we tend to use twice as much water as we need per V60 which is wasteful. Our Simonelli T3 has 14lb of brass in the group head and three temperature stabilisation areas per group ready to supply thermal mass and temp stability without wasting any water other than perhaps the standard company two second flush of the group.

That last sentence reflects how the class started to work. We would flag the problem of water temperature rapidly dropping in a pouring kettle; propose to the group that they increase the quantity of water in the kettle; play upon their conscience environmentally; explain this is possible to get around commercially if you are willing to use a machine. Not just an espresso machine either. A Technivorm can avoid this issue, so can a bulk brewer, but not as well as a T3;). Then in the sage words of Mr Dale Harris we would ask, “is the problem really a problem?”. In the case of a pouring kettle, the downward sloping temperature profile is probably a good thing, like in the case of the lever machine. So the important thing is to fill your kettle to the exact same level each time, let’s say 500ml from a boiled kettle. Then start a timer and commence pouring water after the same displacement of time. Test this with a probe a couple of times. But I put to you, I just got a room-temperature Hario 1.2L pouring kettle and put it on a scale, poured in 500g of just boiled water and straight away took a temp reading that was never higher than 93deg C. Not a problem if your name is Jeff Verellen: 83degC

http://worldaeropresschampionship.com/recipes/

Bit of a problem if your name is Matt Perger: 97DegC.

Hario V60 Pour Over by Matt Perger for St Ali, Sensory Lab and Hario World

To get 97degC in a pouring kettle, it will also need preheating and you need to fill it very full and brew more or less directly. The temp is lost transferring the water and, so you’ll save water using an electric pouring kettle like a Bona Vita.

If you are a home brewer and fill your kettle with the favourite water of @philwbass, the Legendary Tesco Ashbeck Scottish mineral water with 6pH and 80ppm, you’re in front. Commercial coffee-shots can’t compete with this control using an RO system or a water softener. This is one of the few areas in which a home brewer can compete against a barista and win. The water is about 98.6% of the brew. A home brewer can choose their water and preserve it’s integrity with a quick once only splash in the kettle.

The liquid retained in the grinds that doesn’t make it into the cup makes a very significant distinction between brewing’s three species. The liquid retained ratio in your average drip method is about 2g per gram of grinds. The average liquid retained in a coffee shot is about 1.3g per gram. Immersion brews are even worse offenders here at 3g per gram. Here there are implications about lost coffee solids. Anyone who saw Vince Fedele’s Nordic Barista Cup talk last year will remember his suggestion that we must up the dose to compensate for the higher liquid retained ratio in immersion brewing. The liquid added to a drip right at the end of the brewing process is what will be retained. It is likely to constitute a fairly trivial loss of sugars and acids that you might want to drink. Whereas in the average immersion brew, like an Aeropress or a cafetiere, what you lose is the same as what you’re going to drink, so if you’re losing sugars and acids, you might need to up the dose, to get a little more to go around. Prufrock will dose 68g per litre for immersion brews and 60g per litre for drips. Well you don’t have to worry about that with coffee shots. The retained liquid at the end of a coffee shot, just like with a drip, has a fairly trivial TDS and what we lose down the solenoid valve and into the grinds is about half as much as in an immersion brew. We just weighed the exhaust liquid out the solenoid on our T3 and it was 5.2g from an 18g dose. The puck weighed 33.2g so we lost about 20g of water in the process. A cafetiere would have lost 54g approx. A drip would have lost 36g.

Contact time in coffee shots preparation becomes a given because the grinds aren’t fine enough to alter contact time. With an EK43, lower fines production means less risk of blockages from fines migration. So it’s all back to grind size as the principal variable. We can also relax a little about distribution and tamping as channelling will be much less of an issue. Immersion brews are fairly immune to channelling (Aeropress and Clever might be an exception here), but drips might be affected by asymmetric distribution of brew water. Of course all brew methods will benefit from better grind size distribution and home brewers with hand grinders will be at a significant disadvantage where production of fines and boulders are concerned. I used a 1mm sieve with my Japanese hand grinder the other day on my usual Aeropress grind and found I had lost about 4g of my 15g dose.

Unlike a pour over or an Immersion Brew, the grinds are tamped firmly in place in a portafilter. We are able to easily groom them into an even distribution and water will wash over the whole cake simultaneously. Kaminsky talked about the coffee cake as a “static flat bed” and considered this another advantage of using the machine to help keep things stable. Turbulence isn’t a factor, just erosion. The Grinding and Brewing course from the SCAE talks about the three Ts: Time, Temperature and Turbulence (agitation). Your average modern espresso machine has got the three Ts well in hand. Prufrock’s cupping method doesn’t currently involve a big stir down to the bottom of the bowl. We have tested this and find we prefer a standard three-pushes-with-the-back-of-a-spoon kind of break that is easy to replicate. Immersion brewers need to be extremely mechanical with stirring. If you stir a brew to form a whirlpool, it is very difficult to replicate the exact pace of the stir and the level of erosion that will ensue. Cross shaped stirring and a set number of stirs is a decent solution for hand brewing.

To reduce turbulence with a pouring kettle a brewer can bring the spout closer to the slurry, and should consider inserting a flow restrictor, though we don’t use them because we prefer a stronger sluice of water after the bloom to help dispel CO2 more rapidly. We always ask brewers to imagine a string joining the pouring kettle to the ceiling. The vertical distance between the slurry and the spout shouldn’t shift at all to keep flow rates and pouring height equal across the top of the coffee bed. We make sure our hand brews have time and weight markers that are fixed. We prefer the flow rate to be slow and constant to help maintain the same volume of liquid in the cone right up until draw-down. If not slow, then replicable. An espresso machine ain’t so good at slow but it sure is good at replicable flow. The average time taken to fill a 5 fluid oz cup with just hot water from a machine with no restrictors is about 9 seconds, with 0.8mm restrictors it’s about 12 seconds and with 0.6mm restrictors it’s about 17 seconds.

Automation is definitely not a bad word at Prufrock. Gwilym has long since maintained that when a bean-to-cup machine comes along that out strips a manual machine for consistency of extraction yield and taste, we’ll buy one. This day has not yet arrived – as far as we can tell? I have it on good authority that Hoff plans to use the new and not quite yet released Volcano Swift grinder from La Marzocco in Square Mile’s new venture in the Ace Hotel in Shoreditch; the Volcano for milk drinks and an EK43 for Espressos and Coffee Shots. The classic Swift’s mechanism has been modified and put on the much cooler looking Volcano. This will be interesting to watch.

With the issue of speed there really is no contest commercially. It takes us 2 minutes at a walk to make a coffee shot and 5 minutes to make an Aeropress. That might seem crazy slow but we aren’t following the instructions on the box. Are you? 3FE can pull an EK espresso shot in 1min and 16 seconds so this is a reasonable guide. Commercially when multiple brew bar orders come in a customer can very quickly have a 20min wait for a hand brewed coffee. We’re going to keep doing hand brews, we love them but an in-between experience awaits the punter that is time pressured and still wants to have a ‘just for you’ experience.

Texture is an interesting one. People are filtering their coffee shots after brewing. I’m not thinking this is a good idea. The pores in the paper get blocked half way through the draw down.

Lynsey Harley (Q Grader for Falcon) may recall disqualifying me in the 2012 UK brewers cup when I made her a lovely cafetiere and filtered it through a Woodneck cloth (A Wendelhoff) which was taking so long to draw down I just left her with it and legged it off stage.

13.3.3 Competition Time
“…Competitors must actively place each beverage in front of a judge in order to for it to be deemed “served.”
2013 World Brewers Cup Rules and Regulations – Version 2012.10.08 Page 9

Fair enough. Incidentally in this same flight of drinks was a drink I named The Kaminsky, (whipped cafetiere) which I might have inaccurately appropriated from Ben’s US Brewers cup performance.

The appearance of a coffee shot is not great. Much like a Trifecta and to a lesser extent, an Aeropress, they are turbid. There are tiny colloids aplenty, but like with cuppings and cafetieres, this seems to intensify aromas and really isn’t experienced in the palate as grit but rather just increased body. Maybe a tiny bit gritty but it’s not detrimental. Our best coffee shots with the Tanzania are certainly coming after sieving with a 250um sieve but even sieved, you still get colloids. They must break off the larger particles. Or we aren’t sieving hard enough. Here the coffee shot cannot compete with the lovely translucence of a V60.

What about the Mother’s Milk pop-up one of our baristas, Will Hilliard and fantastic James from the Espresso Room are planning on Little Portland Street. It’s the middle of town and they might want to serve pour-overs and Aeropress but in the early days when it’s just two baristas on the rota and they get their first queue and the first guy in the queue says, “I’ll have a syphon please.” The other four people in the queue wait ages and get annoyed and just think, I’m off to Kaffeine. A couple of months down the track they might be able to justify four staff, someone on the brews and dishes, two on the machine, one on the till. This is kind of how the Espresso Room does it but if Mother’s Milk hire four staff at the beginning without investment so that they appear more professional and unflustered when unorthodox orders come in in the early days, they might have to move back in with their parents after the first week. But a pop up or a hatch or a cart, with two baristas and one EK can make fantastic filter-like coffee, espresso and any number of guest options without even moving their feet.

The last issue is the barista’s craft, the artisanal element. I sometimes say in Brew Methods classes, “it’s not a performance, it’s a choreography” but that’s a bit high falutin’. I think Varney has already countered any arguments in this area anyway. Any old excuse to watch it again.


/ / /