I wouldn't consider this SCAE approved.
I was really lucky to be able to attend the SCAE Brewmaster course held at Marco HQ a couple of weeks ago. The course in a nutshell gives someone the tools to effectively take control of their brewed coffee. The aim it would appear, from the SCAE point of view is to raise the quality of brewed coffee across Europe by creating a nucleus of “Brewmasters” to spread the good word. The ultimate target isn’t really the super speciality people, rather the average cafe, who is likely to be using as little as 30g of coffee in per litre water, and extracting every last morsel of coffee from that 30g. Nonetheless the material is hugely relevant to anyone who cares about brewed coffee quality. It was was more than just interesting for me, however, it has crystallised some thoughts in my mind about brewed coffee, particularly in relation to grinders for brewed coffee, which has been a bugbear of mine for some time.
Brewmaster peeps. A great bunch of lads.
Firstly I will say that, in general, in my (good) coffee drinking lifetime I have been drinking updosed underdeveloped coffee as outlined by the SCAE (and most other bodies). For those unfamiliar with these concepts, it basically means I was extracting less than the ideal (18-22%) amount from my coffee, and perhaps compensating for the diminished strength of such an extraction by increasing the dose. To fall within “SCAE Ideal” you need to be using between about 50g /l and just under 70g/l. The Nordic standards permit a higher dose range, the SCAA standard a lower one. With drip and press, I have pretty much always favoured finding myself at the top or just above the SCAE standard (without knowing so at the time). For Vac Pot and Aeropress, however, I have found myself using doses well in excess of any known standards.
However, I am not alone. Examples of underextracted, updosed brews are abound. In the recent Aeropress Championship two of the top 3 methods (including the winner) used a dose of 100g / l. When I duplicated these results at home, I got tasty cups, that were underdeveloped by any standard (12-14%), but fell into the ideal range in terms of strength. Take note that the palates of the judges of the aeropress championship are not to be sniffed at.
So does this suggest a kind of mass hysteria among speciality coffee folks, a laziness in technique due to increasing bean quality, or perhaps that the 50+ year old standards are inadequate?
The answer is probably not 100% any of the above, though I will admit at times finding it difficult to brew a bad cup of certain stellar coffees (hint = Aricha).
An example grind profiles of the same grinder and beans at different settings. Note as the grind size gets smaller, more fines are produced.
The problem though, is the Gold Cup extraction percentages presume a relatively uniform grind, the kind you might get from industrial roller grinders, and to a lesser extent some commercial / bag grinders. The physics of bean grinding, however, says that the finer the grind, the more fines (dust-like tiny particles) will be created, regardless of grinder. Whether you have a Mahlkonig Guatemala or a Rancilio Rocky this holds true. Conversely coarser grinds produce fewer fines.
Over the course of a brew these fines will overextract in relation to the normal sized grinds, typically creating a bitter, muddy taste. I’ve heard it mentioned that the fines add complexity, I would argue brew a cup of 100% fines and see how complex you find it. A turd in a swimming pool probably makes the swimming pool more complex. It seems natural enough therefore that those of us sensitive to these types of flavours , identifying them as undesirable will perhaps instinctively move to coarser grinds that produce fewer fines. When we find the mouthfeel and body lacking we increase the dose.
I’d put 95% of the blame for this trend on the grinds we use.
The other 5% perhaps is that these types of brews actually can be excellent. Punchy, intense, BIG! They may not deliver a full picture of a coffee – but does a ristretto give you everything a bean can give to espresso? It is foolhardly to merely label anything short of 18% extraction “wrong”.
I've been using a tea strainer as a sieve, however, in Japan a device for this very purpose is available. The deviceSTYLE Brunopasso MPS-50. This is on my shopping list.
My message here is not to say that 18-22% extraction is not a meritorious standard. With a good grinder, or as I have been doing at home, sieving my grinds to remove the fines, extractions in this range really sing, beautifully sweet, rounded and complex. Without sieving the grinds there was a kind of muddy overlay to the flavour profile. I do believe this is the default level to which coffee should be brewed. But we don’t all have Mahlkonig Tanzanias or Ditting K804s (yet), and if we do, when were the burrs last changed? Sieving grinds is also a pretty tedious process.
It’s apt to point out that the SCAE Gold Cup is due to undergo a process of consultation to update the standards and aims of the concept. The timescale is in the half a decade kind of region so I suspect one won’t be able to accuse to them of not being thorough. I do hope, however, that in that time we are able to move on from lamenting fines, the grinders that produce them and being drawn inexorably coarser… always coarser.
Great read!
I have so many mixed experiences with the traditional brew charts. Sometimes you overextract and other times underextract in order to get the profile you want.
I just bought a Bunn brewer and in order for the coffee to taste anything near good I had to updose and grind coarser.
What is interesting is that they teach us that most of the flavour (80% or so) is extracted in the beginning of the brewing cycle on a brewer. How big of a deal is the remaining 20% when you know you start extracting shit?
In my opinion, espresso is all about extracting flavour. That is why you stop the process when you see there is nothing more of the goodness coming out of the spout. Why not approach Filter coffee the same way??
Thanks Tim. Interesting points. There is more than one way to skin a cat I suppose. :)
I am concerned my response to this might be a little long winded as there are a few things in the above that I want to discuss further.
I was with you all the way up until:
The gold cup standard has come from decades of coffee brewing, during which time I’d argue that burr design has probably changed quite a lot, as well as what we know about coffee. Interestingly this doesn’t seem to have impacted upon the ideal of 18-22% even extraction.
Fines have a higher surface area, gram for gram, and therefore extract faster. With no fines you’d have to work harder to extract a fuller percentage of the coffee, obviously with too many fines then you’d likely get an overextraction, or at least a portion of the brew being overextracted.
How did your brews chart pre and post sieving? How much coffee are you loosing in the sieving process?
I think fines generally do their worst post extraction if they aren’t sufficiently removed from the brew – that is the most common cause of the muddy cup for me.
Which is perhaps one of the mistakes we make. We are more capable at perceiving strength than we are at correct extraction. We have a fairly ideal strength in mind and we seek the easiest route to get there. I think the fact that most of us don’t understand the nuance of extraction means that the only solution we have that comes to mind easily is the dose. Keeping the grind and increasing the extraction through another method (heat, steep time) would perhaps be the first place to experiment before leaping back to the dose.
I think more interesting coffees are perhaps more tolerant because we can fairly easily get what we want out of them – which is often a strong characteristic. I’ve had lots of cups of coffee that were nice at 17%, but I’ve had lots better since I started hitting the 19% mark.
@Tim
I do. I am trying to get up to an extraction percentage and to go no further. ‘Goodness’ is a very open word in espresso. I suspect you and I might see a little more goodness come out of the spots as the shot starts to lighten than a lot of people pulling massively updosed ristrettos who stop at any sign of a lighter liquid. Again with espresso we are chasing a similar extraction percentage, and colour is a reasonable indicator of how much coffee we’ve taken from the puck and how much remains.
@Tim and David
I think we ought to look at what we particularly like from the updosed and underextracted brews – and seek to retain them but perhaps increase the quality of the cup, increase its quality of mouthfeel and complexity. I think understanding the brewing process better through programs like Gold Cup is a great step in getting there.
There are a lot of good points there, I’ll do my best to respond in a readable fashion.
On the Gold Cup standards presuming relatively uniform grind. Perhaps I should have said that the Gold Cup approach to “rectifying” extraction is linear, whereas grind adjustments are somewhat exponential in terms of the ratio of fines to normal grinds.( ie make the grind finer to get a greater level of extraction).
In other words one 18% is not always the same 18% as the next 18%.
At a fine grind 18% could be 17% from normal sized grinds and 30% from the fines. This may make a relatively small impact on the chart, but it can make a big impact on the palate.
The whole premise of the 18-22% extraction is that below the 18% is underdeveloped and above is overextracted and essentially tastes like shit. Most grinders to a greater or lesser extent are bimodal in their distribution – so there are not 1 but 2 populations of grinds to consider, and inevitably one population gets extracted above 22%.
So when you say, with no fines you’d have a harder job extracting a fuller percentage – I would respond that I don’t want that portion of the fuller percentage. Rather in an ideal world without fines we would just grind finer, until the “normal” population of grinds has an adequately large surface area to reach the desired extraction.
When I sieve I lose about 5%, though this is far from thorough. Visually it is still quite obvious fines remain, but it is an improvement nonetheless.
In terms of brew charts I lose about 0.4 – 0.5% extraction (dose remains the same).
Fines remaining in the cup is another matter entirely. Even with a paper filter, I can get a muddy taste from a grinder that produces a lot of fines. Really I mean a muddy taste, rather than a muddy appearance, or any physical attribute. Perhaps lack of clarity is a better description.
Of course increasing the brew time with a coarse grind would be a valid approach, though maintaining temperature over such a period becomes an increasing concern, and impacts on practicalities.
(PS I cleaned up your blockquote tags – < instead of [ )
Sorry about the blockquote thing – too much time on forums!
I get what you mean about fines, though their presence in many, many good brews I’ve had means that I can’t point the finger of blame at them entirely.
It would be interesting to know more about whether fines extract quicker, or whether they just contribute more solubles due to a larger surface area. Taken to an absolute micro lever – each cell doesn’t know if it is a part of a fine or a larger piece, only whether it is in close proximity to brew water (i.e. on the surface)
Thinking out loud there really – it is late and my brain is a step below functional.
If you take my above numbers to their logical conclusion, the 5% contributng 0.5% more extraction means that population is 10% more extracted.
10% is of course a huge amount in terms of tatse.
I thnk that is evidence that fines extract quicker.
No – it doesn’t relate to the speed of extraction. They contribute more per gram due to an increased surface area, though unless you somehow seperated them post brew and tested them to see if more than 18-22% of their weight was extracted you wouldn’t know exactly what they contributed.
What is speed but the percentage extraction as a function of time?
Let’s assume that my sieving was 100% efficient. It isn’t, but the point is still valid.
In the sieved brew I have 1 population that is x % extracted.
(A) 1*X = 17%
(17% is the real number).
In the unsieved brew I have 2 populations, the normal grind that is X % extracted and the fines that are Y % extracted.
(B) 0.95*X + 0.05Y = 17.5%
(17.5% is the real number)
From (A) we know X = 17%.
Therefore (B) becomes
0.95*17% + 0.05Y = 17.5%
16.15% + 0.05Y = 17.5%
0.05Y = 1.35%
Y = 27%
ie the fines are 27% extracted.
The speed is therefore 27% / 4 minutes = 6.75% per minute etc.
The speed of the “normal” population is 17% / 4 minutes = 4.25% per minute.
Of course the speed will vary over the course of the brew.
Good discussion. This is something that I’ve been spending a lot of time working on, experimenting, and thinking about. I really want the ExtractMojo as a result.
@David: I’d been thinking about trying to sieve-out fines for a few years, but never got around to it. I’m glad to hear your results have been good. The only thing I have to add is that keep in mind that the rate of extraction of any of the particles, including the fines, will vary significantly on the brew method, particularly regarding agitation. Fines play a very different role, depending on the extraction environment.
@James: I’ve been thinking a lot about updosing for brewed coffee, and I’m frankly troubled by it, but for reasons that I’m not myself entirely clear on. I’ve been witnessing doses of 20-60% higher than “standard” lately, and with greater frequency, and it always makes me frown.
What bugs me the most, I suppose, is that it’s not really necessary. I’ve been making outstanding Chemex brews with about 40g for about 0.6L of water, with all of the characteristics we look for. When dosing goes up to overcome the brew method (usually to shorten the brew time), it feels like cheating to me. An inelegant solution to an otherwise classic problem.
The elegant solution: refine your technique or practice your technique. “Refining,” more often than not, seems to be to extend the brew time to a more “proper” duration.
This is merely based on my current experiments and rumination… so it’s a work in progress for sure.
David,
Firstly, I congratulate you on a great article. I have been both captivated by the discussions and delighted to see Gold Cup and brewed coffee being discuused logically.
The Gold Cup is a drive to bring up the standard of filter everywhere. We do not discern, but Speciality coffee is the logical home for Gold Cup brewing. They will envitably find each other through the medium of that section of the coffee community which cares.
I have used the extract Mojo on the Uber Boiler at the weekend in Cologne and while I know the pallet is subjective, a 60-65g/L brew i.e. not dosed higher but extracted between 19-20% was sparkling and balanced. It helped that we used great coffee of course courtesy of Square Mile. These were brewed on Chemex with a Tanzania – as close to single peak grinding I can find). Therefore minimal fines.
I’m all about the science and while my gut tells me a uniform grind with as few fines as possible will deliver the best filter, supporting your argument David, I will aim to scientifically chart the differences in the short term.
Thanks for attending our Gold Cup course and congratulations again on a great article.
My brain is slowly coming around to your point about extraction speed – sorry about this.
@Nick
“Very different” I dunno – somewhat different I accept. It seems logical to say that any brew method that gets a certain percentage extraction out of relatively coarse grinds, will get a higher level of extraction from relatively fine grinds. Whether fines experience a different microenvironment to “normal” grinds under conditions of turbulence seems plausible, though perhaps would only exacerbate the problem. Really entering the world of speculation here so I’ll stop.
@Paul
Yes, sorry. I should have beem more careful on that point.
My understanding of the extract mojo was rammed home this week by Messrs Stack and Kaminsky in Cologne and I am hugely excited about the potential advancements that can be made in specialty coffee using this and similar technologies.
Perhaps the most exciting thing is that by laying down markers in Coffee Brewing (i.e. extraction, dose, water ratios, time) we can begin to ask questions in areas we never had the luxury of asking before.
The ability to describe your coffee in definitive numbers is undoubtedly geeky but an exciting advancement for an industry that has such a strong internet-based community. We now have the ability to define our coffee brewing and communicate findings as opinion based on fact rather than a fact based on an opinion.
[...] Best illustrated (in a not very laboratory but perfecty understandable way)in this great post by DavidWalsh – go to the thread and read David’s coments [...]
so, getting a bit conceptual, why not design a graduated (in size) filter topology, or just some type of grid/topography that somehow is responsive to the migration pattern of fines during the extraction process, thereby segregating the smaller particles – your choice to in- or ex-clude them. alternatively, why not pass off the mechanical grinder problem to an active membrane (filtration) problem – i know it’s a bit conceptual, but this is what membrane scientists do with very complex organic compounds. one of our customers (we call him doctor membrane) does just this in the petroleum field. create a selective membrane, disposable, inexpensive filter, that removes the organoleptics that we deem unfavorable in the end result. i have spent some time discussing this with dr membrance, and, with some expert guidance (extensive knowledge of organic chemistry as regards coffee taste, e.g. dr illy & co), he believes it is doable. sorry, for the diversion – this eliminates the grinder chase which will likely never be capable of removing all fines, unless we come up with a “fines trap” in the exit path.
You sieve your grinds too! I thought I was alone!
It feels so wasteful though that it kind of embarrasses me, especially when the beans are good.
Recently I had the thought that cold-brewing the fines would be putting them to good use. Summer’s coming after all.
I might also be able to use the liqour to spike my hot chocolate with caffeine. Hmm…
[...] 3. http://www.devicestyle.co.jp/en/products/brunopasso/ca6/index.html 4. [...]