Particle size inconsistency and pre-grind blending

#1: Post by **Ron** » July 19th, 2019, 8:29 am

I just saw an interesting video (http://www.youtube.com/watch?v=2lPGf1gM9nA) that shows how inconsistent particle size means that large particles can be very underextracted while smaller particles get completely extracted. So, if you blend two different bean types (like a harder and softer bean or a lighter and darker roasted bean) before grinding, you may wind up with different size particles and then get underextraction in a big percentage of the total.

It would look like you should grind the different beans separately and only blend after grinding. But I'm under that impression that blending is mostly done before grinding and many outlets sell their famous blends in whole bean or even green unroasted form. Doesn't this guarantee inconsistent particle size and fractional underextraction?

What's the solution? What do the pros do?

#2: Post by **C-Antonio** » July 19th, 2019, 11:09 am

there are roasters that prefer blending preroast for that reason, and I suspect green blends are engineered to not pair extremely different beans.

#3: Post by **Jake_G** » July 19th, 2019, 11:12 am

It would seem that the only rational solution is to grind each bean, one at a time at the requisite grind level for each based on roast degree, varietal, moisture content and density. Then of course you would need to sieve the results and reconstitute at the optimum ratios to achieve the optimal puck composition and stratification.

#4: Post by **ojt** » July 19th, 2019, 11:47 am

I have the same problem. In fact, based on my very limited experience, it is better to use very simple blends or just single origin. Or just enjoy the coffee and not worry about it too much. I have one blend here that I keep finding too complex for my taste. It's sort of like what you describe; over extracted for one part, under extracted elsewhere, etc. It's by no means a bad coffee, but I prefer simpler and clearer tastes.

July 19th, 2019, 1:53 pm

I think you can approach this by grinding the beans separately on different grinder settings and trying to get the same degree of fineness.Then combine the grinds. Up to now I've been been grinding the two together which is not optimal.

#6: Post by **happycat** » July 19th, 2019, 2:12 pm

Ron wrote:I just saw an interesting video (http://www.youtube.com/watch?v=2lPGf1gM9nA) that shows how inconsistent particle size means that large particles can be very underextracted while smaller particles get completely extracted. So, if you blend two different bean types (like a harder and softer bean or a lighter and darker roasted bean) before grinding, you may wind up with different size particles and then get underextraction in a big percentage of the total.

It would look like you should grind the different beans separately and only blend after grinding. But I'm under that impression that blending is mostly done before grinding and many outlets sell their famous blends in whole bean or even green unroasted form. Doesn't this guarantee inconsistent particle size and fractional underextraction?

What's the solution? What do the pros do?

I think it's problematic to look at only one factor within the entire coffee system.

The entire system includes inputs, processing, and results.

Based on our tasted results, we adjust our inputs and processing. So the fact that we may have inconsistencies in the grind or roast of two inputs can become irrelevant if we've adjusted our blend and processing (grind and temperature and timing) to achieve our desired tastes. This assumes you are using decent equipment.

You can prove this for yourself. Sift your coffee to remove all the boulders or all the fines and then maintain your processing and consider your results. For me, sieving coffee tended to make it worse. I would have to start from scratch in reformulating grind, temperature and timing. Considering how much coffee is wasted in sieving, I couldn't see the point.

I will admit that a crap grinder produces a gravely distorted coffee system. For instance, an entry-level Encore produces too many fines within a medium grind. The result is horribly muddy flavours and a temptation to keep making the grind finer and finer to make it more uniform, with the result that you need cooler water, smaller doses, and faster brew times to have a reasonable drink (albeit one that can become unidimensional).

If you check out a video on Illy, you can see how they use sniffers/tasters to review a variety of input coffees (which change in taste and aroma from season to season) to determine what blend ratios are required "now" in order to produce the branded flavour they are known for.

#7: Post by **jpender** » July 19th, 2019, 2:17 pm

An interesting lecture that could have been given in under 15 minutes instead of 80. But somehow, on one of the main points, that Justin Timerlake guy who never introduced himself failed to give a cogent explanation for why he was "pretty sure" that overextraction doesn't exist except under extreme circumstances (e.g. boiling coffee in an ibrik for several minutes).

My takeaway was that unless my coffee comes from a roaster who uses a fluid bed roaster and is ground in an EK43 that I will inevitably end up with something hardly worth drinking.

#8: Post by **Jake_G** » July 19th, 2019, 3:49 pm

So I actually watched this after my cheeky reply above (which I still stand by) and Matt has a point, but his pseudo science really bothers me.

"Let's design an experiment that gives us predictable results and then interpret the results to confirm what we already knew and call the process "science the hell out of it" so there can be no question that we're right."

So here's the 15 minute version of the 80 minute talk condensed into cliff notes:

Part one:
Water extracts the part of coffee that it can touch easily.
Smaller particles have a higher surface to volume ratio and extract more readily than larger particles.
Wider spreads of particle sizes lead to a wider spread of extraction.
Tighter spreads of particle sizes leads to tighter spread of extraction.

Part 2:
Higher extraction yield tastes better than lower extraction yield until it doesn't.
Find the highest extraction before it tastes bad again and extract there.

Yeah. 80 minutes is a bit much to cover that, but ok.

I covered this a while ago on why I thought that more unimodal burrs taste better at higher extraction yields than less unimodal burrs. I don't think he's wrong in that coffee will taste better the more you extract until it starts tasting bad. The problem is when you take this and assume that you can apply it outside of the narrow confines of where the observations were made.

Case in point:
"EK43 extracted at 22% and tasted great.
Robur extracted at 19% and tasted bad."

Great. On the flip side, if you extract a unimodal flat burr set at 19% it tastes awful. And if you pull a Robur shot psycho lungo and get 23% (science says this is possible) out of it it also tastes awful.

All that tells you is what it tells you. Nothing more. You can't make sweeping generalizations that roburs are bad because they don't behave like an EK43. You also can't say that EK43s are bad because they taste bad when you extract at 19%. All you can say is that different grinders taste different and if you prefer the taste of one over the other, cool.

Some coffees will taste better and better as you tighten up the grind and soften the puck with more and more preinfusion. Some hit a limit and need to be pulled shorter to taste best. I tend to prefer single origin micro lots where I can focus on what's going on with a bean because that's part of the fun for me. But I also enjoy blends and really don't worry about the fact that some beans may be (will be) more or less extracted than others. You find the relative maximum where flavor is best and roll with it.

I will concede that generally I find that I prefer shots that are more extracted and I certainly don't find myself lamenting that a shot would have been tastier if only it had some channeling to "liven things up a bit". But the pursuit of refractometer bliss is a bit tiring to listen to for 80 minutes.

Cheers!

- Jake

#9: Post by **Denis** » July 19th, 2019, 4:03 pm

Nope, long shots with long PI will always taste inferior to short shots with high flow in the cup, at the same EY.

The long PI adds a bad aftertaste in almost all cups, and I know that because I did that for more than 1 year, on the little brother of the Robur that is the Kony E, and now I am pulling on the grandfather of the ek43, the Guatemala SB.

Soaking with long PI allows you for better naked extractions, but this asks for a finer grind than a medium time PI. The finer grind is not the solutions, as Matt tries to tell in that video, that I posted 3 times already in the last 6 months here. He is saying finer grind means a bigger total area to extract from, and smaller diameter particles means a better extraction overall, because water penetrates only 100 microns inside. This means the Ideal particle to be saturated and penetrated with water needs to be 200 microns. I grind ed with a peak at 200microns and with a 2-minute shot time. The taste was bad, and its like that because of an uneven extraction, the while the first part and the medium part are ok, the aftertaste just destroys it.

Every time I pull a coffee, when I go fine with it and do either long PI, lower dose, bloom shot it tastes worst and the EY is always lower, no matter what. The solution is not fine grind, because with fine grind, the water will now flow with ease in the whole puck volume, but it will find spots where it will be choked, because of the distribution, or some crowded filled with fines spot that stick to the bigger chunks, so because of those it will find a way out, and that is just channeling, or uneven flow, or dry spots on naked PF on the last half, or the water is forced to flow at the edge of the basket wall.

I always believed finer you can pull, the better. It's not the case anymore, this applies to all coffee roast range (light-medium- dark- 2nd crack coffee- Italian south roast, or Vienna roast.

And a small hint: a soak long PI shot, pulls back the need of WDT. After getting used with coffee at 23-25-27 % EY when you taste a shot that is 18-20% its just....bad.

#10: Post by **Jake_G** » July 19th, 2019, 4:25 pm

I appreciate your perspective but will always struggle with comments that include the word always

.

I think the trouble with EY is that it's nothing more than fuel economy for brewing. How much stuff you extract doesn't really correlate in any meaningful way to which stuff you extract. 18g, 25% EY means that 4.5g of coffee solids made it into your cup. That's all. Was that 4.5g the best tasting 4.5g that were available? Is there any way to know? If I got another .5g out of that same 18g what would convince me that the 5g of coffee solids would somehow be better than the 4.5g I had in there before? Furthermore, how do I know that 4.5g is better than only 4? It all depends. I've seen no convincing evidence that more is better. It's just measurably more and the FlavorTron 5000 flavor analyzer hasn't yet been released.

My preinfusion is actually shorter these days, but I'm one of those odd ducks that likes body and mouthfeel as much as anything else, so I tend to like a more concentrated shot. Even if something tastes "more evenly extracted" by going with a higher brew ratio, I find that I don't often prefer the experience of drinking it. So I tend to go for a shorter shot in terms of yield because I prefer it. I don't really care what the EY is because I've noticed that I prefer tasting what I extract over testing it. I'm a nerd and I love data and really enjoy the process and I pay attention to things and look for trends, but I'm just not hung up on chasing anything other than what tastes good in my mouth. You know, when I taste it. With my tongue.

I think all that Matt's video can claim to demonstrate with the cupping exercise and the 100 micron extraction observation is that you can maximize your fuel economy by pulling shots with an average particle "size" of 200 microns. To your point, if doing so doesn't taste good, who cares? Well, if the specialty coffee industry cronies (I'm not claiming to know who may or may not fall into this category) can convince their clientele that espresso brewed in this range is the only thing worth drinking, then they can maximize their profits. Good for them. I'll keep brewing under-extracted (or I guess unevenly-extracted) swill at home.

I hope I'm not coming off as too argumentative here. I just get annoyed with sweeping generalizations that don't appear to be based on anything useful. I've yet to see blind taste results where higher EY unequivocally dominated in terms of flavor. When i do, I'll eat crow and joint the masses.

Cheers!

- Jake