Skip to content
Coffee Ad Astra
Coffee Ad Astra

Coffee Ad Astra

Coffee Brewing with a Scientific Focus – Jonathan Gagné

  • Home
  • The Blog Posts
  • Tools
    • Grind Size Application
    • Flavor Wheels
    • Brew Water
      • Bottled Water Spreadsheet
      • Water Recipes Spreadsheet
    • Coffee Filters Analysis
    • Recipes
      • V60 Recipe
      • A V60 Pour Over Video
      • Siphon Recipe
    • Coffee Control Chart
      • Percolation
      • Immersion
    • Cupping Statistics
    • Coffee Glossary by Mitch Hale
  • Astrophysics
  • Instagram
  • Patreon
  • Contact

Category: Data Analysis

Posted on August 4, 2019

An In-Depth Analysis of Coffee Filters

In this longer than usual post, I do a very detailed microscope analysis of various V60, chemex and siphon paper and cloth filters.

Posted on July 23, 2019July 25, 2019

How Coffee Varietals and Processing Affect Taste

In this post, I used tasting notes from 1500 bags of coffee too bring out how perceived flavor is affected by coffee varietal and processing

Resources

  • Grind Size Application
  • Current V60 Recipe
  • A V60 Pour Over Video
  • Current Siphon Recipe
  • Flavor Wheels
  • Water Recipes Spreadsheet
  • Bottled Water Spreadsheet
  • Coffee Filters Analysis
  • Cupping Statistics Widget
  • Extraction yield equations
  • Coffee control chart (percolation)
  • Coffee control chart (immersion)
  • Coffee Glossary by Mitch Hale
  • Specialty Coffee Association Standards
  • Barista Hustle
  • Scott Rao’s Blog
  • Mitch Hale’s blog

Pages

  • Coffee + Science = (°◡°)
  • Contact
  • The Blog Posts

Search this Website

Categories

  • Announcements
  • Blind Tasting
  • Brew Methods
  • Coffee Filters
  • Coffee Grinders
  • Coffee Processing
  • Coffee Varietals
  • Collaborative Posts
  • Cupping
  • Data Analysis
  • Espresso
  • Experiments
  • Extraction
  • Maths
  • Particle Size Distribution
  • Perception of Taste
  • Physics
  • Recipes
  • Siphon Brews
  • Storage and Maintenance
  • Tasting Notes
  • Tools
  • V60 Brews
  • Water for Coffee

Instagram

Is this placebo or the consequence of a material sensory phenomenon ? I have no idea, but the same coffee does appear different in the two @kruveinc glasses 😁 Today I’m testing them out with a very weird and enjoyable coffee ! @father_coffee generously sent me a carbonic macerated + washed Kenyan, their Amber 074, which I brewed as a V60 with my “classical” method an Rao/Perger water (45 sec bloom; see link in bio for all details on my blog). I got a 323g beverage at 1.51% TDS (22g dose) for an average extraction yield of 22.2%. The result is a deep, sweet and intense cup with strong notes of blackberry and a winey finish that I really enjoy. As it cooled, I noticed notes of caramel and blueberry too. For me, the Kruve glass on the left emphasized the juicy and sweet side of it; and the one on the right emphasized a bright acidity and more floral notes. Not exactly a controlled experiment or a blind tasting, but it’s interesting! One of my worries was that they’d be too big because I got used to drink my V60s in very small espresso glasses, but I must say I enjoy their size, it’s a bit less than a mug. I could split my 324g beverage in both and I had about two espresso glasses of coffee left. I have been excited to try these glasses for several months now since I backed their kickstarter project, so I’m happy that they live up to my expectations 😀 #coffeegeeks #pourover #kruve #coffeeadastra #coffee
Who wants more discussion of bloom duration ? 🤓 Today I tried preparing two V60s with a 3 minutes bloom on different coffee beans, and both tasted much worse and showed only a slight ~0.2% extraction yield (EY) increase over my usual 45 seconds bloom ! The first brew was Andy’s @canadianroastingsociety Nano Challa Ethiopian (roasted this month), it’s a very well developed roast that tastes great w/o any astringency at ~22.5-23.0% EY. The long bloom had 22.8% EY and tasted astringent, muddy and slightly bitter. I think I may have caused some clogging which could maybe explain the worse taste. The second brew was @ghowellcoffee’s Tabe Burka Ethiopian (roasted in February and vac sealed since then). I usually get ~21% EY with it and absolutely love its floral, lime and chocolate taste. But this time it tasted bitter, ashy and muddy ! For this brew I don’t see what else could have caused the worse taste. I’ve heard a few similar negative experiences with the long bloom so far (@aryanjoshani ) and some friends have had a mix of great improvement or a more mitigated improvement by using it. This tells me using a longer bloom probably only works great on *some* coffee beans ! I now suspect the lighter roasts come out better with a longer bloom, which would make it a bit more likely that the long bloom’s benefits has something to do with coffee cellulose being too stiff in lighter roasts. We need more data, but that would be incredibly useful if we could tweak the bloom when we find a roast is too light for our taste 😮 #coffeeadastra #coffeescience #coffeegeeks #pourover #longassbloom
Today I decided to gather more evidence on how @umeshiso_'s long-ass bloom affects extraction yield (EY). I brewed the same coffee as yesterday, with a 45 sec and then a 3 min blooms. I added short water pulses during the 3 min bloom so that coffee particles absorb all the water they can during the bloom. Again, the long bloom caused an increase in TDS from 1.44% (21.4% EY) to 1.53% (22.3% EY) and a sweeter taste ! The first two images show my brew curves; they put to rest my confusion that yesterday’s long-bloom brew was slower post-bloom. This was caused by my waiting for each pour to completely drawdown 😅 Now both brews have a similar V60 flow rate (~2.5 g/s) and bloom-corrected drawdown times within 4 seconds of each other. In the next image, you'll find my last brew's best-modeled liquid retained ratio (LRR) increase during brew (see my earlier few posts for more). This time I did not wait for the drawdown between each pour. The LRR increased more slowly (blue line) than the short bloom case (red line), but this time it allowed me to reach a higher LRR of 98% of its final value by the end of the bloom (blue circle), rather than just 60% (red circle). In other words, I *succeeded* to get water absorbed as deep in the coffee particles as it will get, yet I saw zero benefits on EY and taste compared to yesterday’s long bloom. This may indicate these very deep layers that get wet last don't have time to participate much in the extraction anyway (I think deeper layers extract exponentially slower). In the fourth image, I show the fraction of light reflected at different wavelengths (colors) on my long-bloom brew (diluted to 1.43% TDS) compared to that of my short-bloom brew. The brews have different colors, indicating different chemical compositions: the taste differences don't seem to be a collective illusion 😬 In the last image I show two other brews prepared similarly; they look exactly alike, so the reflectometer data seems replicable. Currently I think coffee compounds start diffusing into the interstitial liquid during the long bloom, and that's why EY is higher - thanks @samosmrke for the idea !
I just had a look at my Acaia blueprint from this morning's 3 minutes bloom V60 (see my previous Instagram post for more details) and water absorption by coffee particles does *not* explain the different taste and higher extraction yield ! Above, I show in black the water weight in my V60 vs brew time, and in red the expected floor in between pours from the rate of water absorption that I calculated with a normal bloom. As you can see, the valleys between pours do *not* align with it ! I made sure water had drawn down completely between every pour, so they should. You'll notice that I made a small 22 grams pour right after bloom; I did so because if interstitial water was absorbed by coffee particles, I may be able to replace it with this small pour and measure a valley that hits the red curve better. In the second image, I cut this morning's brew data between 45 seconds and 3 minutes, to see whether water absorption was happening at all from the interstitial liquid, and it matches up *very* well ! This means water absorption stopped happening as soon as water passed through the coffee bed, as @apaxos suspected. Another puzzling thing is that my brew lasted for not just 2 minutes 15 sec longer than usual, but instead it lasted for another additional minute and a half, which I could only attribute to a slower flow of water through the coffee bed. I'll be very surprised if that was due to water being that much cooler. These results are puzzling to me, here are my take-outs for now:
(1) Maybe this morning's brew was a fluke, and had a high EY because of a happy mistake I made ? (2) Maybe another mechanism that I don't understand makes the longer blooms different with higher EYs and sweeter taste ! (3) We might be able to improve things further with a non-agitated 3 minutes immersion-style bloom where water absorption would still be happening during the full bloom. Tomorrow I'll prepare the same coffee with my traditional method without a temperature probe in it (that created bypass water in my "traditional" pour over from yesterday), and then I can compare how permeable the coffee bed was, and whether I still get such a high EY.
This morning I just tried @umeshiso_’s long-ass bloom based on brew data I collected yesterday that seemed to indicate coffee particles needed more than 45 seconds to absorb their own weight in water (see my previous Instagram post for more details). I have not yet compiled my Acaia brewprints to measure LRR after the bloom, but the sensorial and refractometer results are already beyond my expectations. I brewed a Burundi coffee roasted by Andy Kires @canadianroastingsociety in May (vacuum sealed since them), and even recent brews I made with this coffee with my normal recipe were almost a full % lower in extraction yield ! 😨 With a 1:17 V60 and my EG-1 grinder at 7.0@700 RPM (burrs touch at 1.6), I usually had an extraction yield in the range 21.0-21.6% (this is one of Andy’s lighter roasted batches of the Burundi). With a 3 minutes bloom I obtained a 22.6% extraction yield with the percolation formula ! That can’t just be aging because I had much lower EYs just days ago. The coffee tasted much sweeter and rich than usual, without any signs of astringency, bitterness or sourness. I recognized the complex profile of acidity and aromas of this delicious coffee, but the whole flavor profile seemed much deeper and more enjoyable. When I compile my brewprints, we’ll see if I actually helped LRR go higher by the end of bloom, or whether this is a fortunate finding that can be explained by something else. I used my cork lid to keep the bloom as warm as possible, but I don’t know yet if that was useful. I don’t want to take credit for this long bloom method. @umeshiso_, @tds_times_brew_mass_over_dose and @apaxos used it way before me, although I had never heard quantitative assessments of its effect ! Oh, and I just received my @kruveinc glasswares. So far I like them a lot ! #coffeeadastra #coffeescience #coffeegeeks #pourover
Are we not blooming our V60s long enough ? Thanks to my Patreon supporters, I just received a second Acaia scale that I will be using to log the water in my coffee pot as well as the water weight out of my kettle in real time. I'm hoping that this will allow me to do a lot of interesting things, one of which to estimate a live coffee bed permeability during a brew and identify channeling, fines migration or erosion as it changes during the brew. What I really did not expect is to already discover something super interesting after a single brew with the two scales. In the figure above, the line in black above shows the difference in weights measured by the two scales; it represents how much water is entrapped in the V60 as I prepare my brew. This is one of my typical two-pours V60 brews (for an example video of my V60 brews, see link in bio for my blog). At the end of the bloom (45 seconds), there is never any column of water above my bed of coffee, and the same is true at the end of the brew. At the end of my first pour, I always wait until there's just a small amount of water above my coffee bed. You might know that a typical V60 coffee bed retains about twice its mass in water weight; approximately half of this is water retained between the coffee particles, and the other half is water absorbed inside of the coffee particles. We call the ratio of retained water over the mass of coffee the "liquid retained ratio" (LRR). My brew had LRR ~ 2.2. What shocked me here is that by the end of the bloom, the LRR at that moment is FAR below 2.2 ! In fact, it seems like the LRR is an asymptotic curve with a MUCH longer characteristic time than I expected. At the end of my bloom, LRR is approx. 1.3, and LRR ~ 2.0 is only reached after 2 minutes ! This could mean coffee grounds are still absorbing water after 2 min, and deep coffee cells are gradually being unlocked for coffee extraction as water reaches them. If I'm right about this, we may gain a significant amount of extraction and evenness of extraction if we did much longer blooms of 2-3 minutes. This could however create new problems (e.g., low slurry temperature ?), so let's try it out and see what happens !
  • RSS - Posts

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 197 other followers

Blog Stats

  • 93,030 hits

Credits

Website logo by Daniel Demay

All the better-looking photos on this website are by Noé Aubin-Cadot

Resources

  • Grind Size Application
  • Current V60 Recipe
  • A V60 Pour Over Video
  • Current Siphon Recipe
  • Flavor Wheels
  • Water Recipes Spreadsheet
  • Bottled Water Spreadsheet
  • Coffee Filters Analysis
  • Cupping Statistics Widget
  • Extraction yield equations
  • Coffee control chart (percolation)
  • Coffee control chart (immersion)
  • Coffee Glossary by Mitch Hale
  • Specialty Coffee Association Standards
  • Barista Hustle
  • Scott Rao’s Blog
  • Mitch Hale’s blog
  • RSS - Posts

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 197 other followers

Blog Stats

  • 93,030 hits

Category Cloud

Announcements Blind Tasting Brew Methods Coffee Filters Coffee Grinders Coffee Processing Coffee Varietals Collaborative Posts Cupping Data Analysis Espresso Experiments Extraction Maths Particle Size Distribution Perception of Taste Physics Recipes Siphon Brews Storage and Maintenance Tasting Notes Tools V60 Brews Water for Coffee
WordPress.com.
Cancel