Note that Lance Hedrick gets credit for recommending this technique. Lance was doing WWDT almost a year before this technique started getting talked about by other coffee content creators. This post also leans heavily on the concept of fines migration as described by Jonathan Gagné in his book “The Physics of Filter Coffee”.

Some of the best discussions on filter coffee brewing originate on Telegram coffee chat groups. Back in March, Steve Bercovici pointed out that on his Ode fitted with SSP 64mm Multipurpose burrs (henceforth called 64MP), he wasn’t being able to hit reasonable extraction yield (EY) targets using techniques that have worked for him on other burrsets. Although this sounded a little odd, I was soon able to replicate this behavior on my Frankenjolly fitted with the same burrs, wherein in the 600-750 um target burr gap range I struggled to generate enough contact time to facilitate thorough extraction. Imagine 3-4 minute brews on a tricolate for 20g doses at 1:16 ratios (brew duration is including a minimum of 90 second blooms). This was an ironic moment of not being able to “choke” the filter.

This seemed at odds with (and kinda still does) the rave reviews that folks were handing out for brews on 64MP burrs. While I do not doubt the sensory experience of such remarks, I seldom saw them accompanied with an EY number for context. To me, the only way to get past this hurdle seemed to be to grind significantly finer at around the 450 um target burr gap for tricolate and use a ratio of 1:11 (or alternatively 375 um burr gap and 1:12 ratio on a V60), using a bloom + two pour slow flow rate continuous kettle stream pour. So while the high-tds brews (2-3.5 tds depending on coffee and brewer) were tasty, and to some extent could bring back the comfort of low-tds brews by adding back some brew-water for dilution, it bugged me no end that I wasn’t able to do this with ease at more coarse grind settings. I will come back to the importance of why coarse grind settings matter.

Have you tried needling the brew

I was ranting on my Instagram stories about how 64MP burrs could potentially be generating less fines than most burrs we’re used to and which is why fine grind + low ratio might be the way forward, when Lance Hedrick reached out saying (and I paraphrase) “Have you tried WDT-ing the slurry?” I thought at first he meant just the bloom and he was like “No, no, the slurry itself. You start by gently scraping the bottom and make your way to the top of the bed”.

This made a lot of sense. Instead of relying on the limited capabilities of a kettle stream (or an extremely uneven shower stream in case of tricolate) to cause agitation, and by extension fines migration, one can use a WDT tool to evenly dislodge fines from the coffee grounds and have them reach the filter.

Remember that when one talks about agitating the slurry, it’s a double edged sword – in brewers with bypass they enable better mixing of grounds with fresh water from the kettle (something conical brewers are in dire need of for the bottom of the bed), but they also dislodge fines in the process, usually proportional to the extent of agitation. When one uses the term “I choked the brew”, what this usually implies is that they caused much higher fines migration into filter pores than was necessary, thus causing slower percolation of water and increased contact time. This is why for example one should consider using a higher kettle pour rate (and pouring lower) for coffees with higher fines generation at the same grind size.

The problem in this case is that kettle stream agitation from the slowest and tallest of kettle streams isn’t sufficient to cause enough fines migration into filter pores to slow down the water column sufficiently.

So you’re saying it works?

It’s quite amazing how well the concept works in practice. For brews that I had a hard time pushing beyond 4 minutes on a tricolate at 700 um burr gap, with WWDT I was now able to push beyond 12 minutes for 850+ micron burr gap for the same coffee. I now use this technique not only for 64MP burrs but even for burrs that seemingly generate higher fines like 1ZPresso KPro. With increased control on the extent of fines migration, one can keep drawdown time the same over a pretty wide grind size range and thus dial-in grind without having to worry about how much it affects contact time.

There’s a couple of things one has to keep in mind though. How much WWDT a brew needs is dependent on coffee+burrs+brewer type. For example on a tricolate, doing sufficient agitation during the bloom can cause enough migration for the rest of the brew with KPro burrs, to not require any WWDT at all for the remaining pours. For a low fines generating coffee and 64MP burrs, one may need WWDT for the bloom and subsequent pours. Other brews may need something in between, like for example WWDT for bloom and not for first pour but again for second pour because the first drawdown happened too fast.

There’s another peculiar behavior you may end up noticing with WWDT – sometimes for the same amount of agitation, a coarser grind size may end up with a longer drawdown time. I thought I was noting the wrong values when Jonathan Gagné confirmed he has noticed similar behavior, and speculated that it could be related to the coarser grind leading to increased bed porosity, i.e. fines will have an easier time traveling through the coffee bed. I have however only seen this happen at certain grind size ranges for certain coffees.

While one may think that it might be to our advantage to have as much fines migrate as possible for a super-coarse grind, remember that if your drawdown is too long, your slurry temperature gets impacted, and the additional contact time may not lead to higher EY due to insufficient temperature. You’d be surprised just how quickly slurry temperature plummets even in something with low thermal conductivity like a tricolate. Although this should not discourage you from letting your brews last longer than 10 minutes on a tricolate, I have found that going beyond 12 minutes (based on the recipe I describe further down) becomes a point of diminishing returns for the grind size ranges we’re exploring.

River rocks

The motivation to use super-coarse grind sizes stems from both the possibility that finer grind sizes can lead to more astringency (check out Gagné’s fantastic blog post on the concept here), and my pocket-science working theory that the finer end of a boulder distribution tends to be extracted more thoroughly and shifting it right (coarser) and extracting them less efficiently leads to less chances of muted drinks. I have found that at very fine grind sizes, using traditional brewing techniques one has to walk a razor thin line when trying to achieve higher EY to not end up muting the brew with just a slight slip up in the high-EY direction. Using super-coarse grind sizes however to achieve similar EY targets in my experience allows more room for error when brewing, even when pushing EY. And my hunch is that even though you’re likely sacrificing some solubility of larger boulders, the slightly reduced extraction of smaller boulders has a larger impact on taste and flavor separation.

Note that I’m not talking about fines here but rather “finer boulders”. I’m not convinced that fines contribute much in terms of measurable EY. Fines however tend to have a massive impact on how water percolates in both espresso and filter brewing, and therefore indirectly contribute to EY.

If you were to ask me how coarse one should grind to effectively use WWDT, I’d like to point in the direction of EAF discord and say “river rocks” size, i.e., don’t be afraid to grind so coarse that the grounds start (metaphorically) resembling pebbles. It’s not uncommon for me to grind in the 750 to 900 um burr gap range with 64MP burrs and between 8 and 0 (one full turn coarsefrom burr touch) marks on a KPro. Coarser than this I tend to find that fines migrate a bit too easily. (If you know who first used the term “river rocks”, please let me know so I can credit them.)

The Recipe

For a short video tutorial on the recipe , feel free to check out this video on Instagram, or alternatively those who prefer Alphabet over Meta, here on Youtube.

Here’s a baseline recipe to use WWDT effectively:

• Start with a dose that’ll ensure both sufficient bed depth and enough fines generation for the filter area the bed will occupy (I prefer a minimum of 20g for both tricolate and V60).
• Pour 4-5x by weight of your dose as bloom water and stir the blooming slurry aggressively with your preferred WDT tool. I have found a minimum of 80g of water to create a slurry that stays fluid enough for sufficient agitation if you need few extra seconds of WWDT before it begins to clump.
• For flat-bottom brewers like tricolate, I prefer using WDT tools with wider needle spread like this one. For conical brewers like V60 I’d recommend going with a narrow needle spread like in this one, to access deeper parts of the bed without damaging the filter.
• Let it bloom for a minimum of 90 seconds, or preferably even longer (particles in coarse grind range especially will benefit from this)
• For the first pour I like pouring upto 65% of total brew-water by weight. Use your WDT tool to agitate the slurry. Conical brewers will inherently need more aggressive WWDT to cause sufficient fines migration due to the larger filter area under the coffee bed. This also means that you’ll need to be more careful during WWDT so as to not tear the filter. Tricolate on the other hand is more susceptible to paper lift rather than tearing.
• Do a Hoffmann Swirl/Rao Spin/Hedrick Shake to flatten the bed.
• When the slurry has drawn down to about 0.5 to 1 cm above the bed, pour the rest of the brew-water.
• Whether or not you need WWDT for the second pour (or for that matter even the first) will depend on your coffee + grinder + brewer combo. For example on a tricolate, with a 1Zpresso KPro, a given coffee may only need WWDT during the bloom to generate sufficient contact time even at river rocks grind sizes, both due to the brewer having very little bypass and the filter area being relatively small and therefore easier to clog intentionally, besides the burrs generating more fines than say 64MP. The same coffee on a V60 despite a slightly finer grind than a tricolate may need WWDT for bloom as well as during both pours to ensure sufficient fines migration and facilitate appropriate contact time. With 64MP burrs the same coffee may need WWDT for all three stages in a tricolate.
• Do a Hoffmann Swirl/Rao Spin/Hedrick Shake to flatten the bed.
• I usually prefer a ratio of 1:15 dose:brew-water by weight on the tricolate and 1:17 on V60 as a starting point and adjust based on taste.
• Always start out sufficiently coarse and then go finer to waste less beans. Low EY brews in my experience are more palatable than high-EY mud water.

TL;DR

• SSP MP 64 mm burrs could not choke the filter sufficiently at coarse grind sizes.
• Lance Hedrick recommended using WDT to agitate the slurry after pouring (not just bloom) to cause intentional fines migration
• This made high-EY extraction possible at super-coarse grind sizes due to increased contact time.
• This approach allows having similar drawdown times for a wide grind size range
• Extent of WWDT will be determined by coffee + grinder + brewer combo
• Grinding super-coarse reduces both potential for astringency and muted brews, besides also allowing a wider margin of error (with WWDT enabling higher EY with contact time).

If you’ve made it till here, consider supporting NOT me, but the hourly coffee worker by donating to GoFundBean

Subscribe to keep your pocket updated with coffee science