Water Retention in Pour Over

With pour over methods we know that a certain amount of water is being withheld (absorbed) by the bed of coffee grounds above the filter.
Should we be compensating for this when brewing and adding water to have a true final weight and strength of brew in our vessel? Otherwise the recipes we write are not actually the true brew if we don’t compensate. Is this correct?

I agree!
Say if we are aiming a coffee with 1.25% TDS from 20% extraction,
then the ground/liquid ratio should be 1:16.


Only measure yield when you are brewing a filter coffee with a stand is kind of mess,
Normally we will cut the brewing by moving away the brewer or the decander( or cup),
but no matter witch you do, you will get something wet.

Always good to know how many liquid we got and then do some ajustments, but I think we don’t need to do it on every cup because of the inconvenient.

Drip brew ratios allow for the fact that some liquid will be retained in the bed, so you don’t need to add additional water unless you’re aiming for bizarrely tight tolerances in concentration (I don’t know why anyone would do this).

Different brewers & even different grind settings in the same brewer, can change the amount of retained liquid, but even if the beverage weight varies across brewers, they are all still the “true brew”. You might see +/-0.05% difference in concentration at the same extraction, which isn’t much. I guess you could brew at say 58g/L for Melitta & Kalita brewers, then 62g/L for bigger Chemex brews to compensate, if you really wanted.

I tend to use a brew stand on scales for smaller drip brews, then I have the cup on a second set of scales on the brew stand. This means that for a given recipe I can ensure a consistent beverage weight out, relative to the water added. If I keep the brew water to +/-1g then I can keep the beverage to +/-2g easily enough. If you have a cup/server of a known weight (e.g. my Chemex is 586g), you can also weigh this at end of brew, after dumping filter & grounds, & deduct vessel weight

Always let the brewer drip out for a little while after the brew appears to have ended, this will help consistency from cup to cup. If you’re pulling the brewer with +/-5g in the cup for 200g average brew then that’s a 1% extraction span, even if you get everything else perfectly consistent…which you won’t.

Different filters also have an effect, given same dripper and grind settings.

Gino filters have a 20% lower flow rate over the same size Kalita filters. This allows longer brew times with much lower coffee to water ratios, particularly helpful for coffee beans that are “fast”, lower elevation or particularly dark roasts.

There really is no such thing as to what should be the grind/liquid ratio, each coffee will express itself differently at optimum ratios given flow rate, grind size, water temp, water formula, coffee(elevation/country/type/roast/…) And there are multiple optimum ratios for each coffee, there is no one “best”…

It is why, no matter how awesome a super-automatic will be, expediency for a good cup will always miss the other optimum settings that provide a different experience.

“Science” while extremely helpful, will never supplant “Art” for achieving the optimum cup except… though an ecosystem that allows brew recipes to be directly shared gets close. Then… we can have multiple recipes per coffee. Of course, at the edge, this is limited as well unless the brewing method is variable such as dialing it in from immersion to like a higher flow rate of a v60, and everything in between.

I digressed…

Back to filters, I have found every so often, a Kalita filter where the flow rate is “fast” so there is variability even among the same filters. However, we usually blame anything else but the filter. I have measured it directly.

Long story, just to say I never pay attention to brew weight :smiley:


Interested into how you tested it? did you use just one brewing device to test all the filters, or did you test the filters in their respective brewer?

As for super-automation, if one of the machines is sophisticated enough all you need is a coffee person with a good palate to set the recipes on it, and then it won’t deviate as much as a human can until something breaks/manually changes. So in theory a super-auto could make an espresso with the same TDS and EY% more consistently than a human can, There was a super-auto at this years WOC that had flat burrs and essentially a puqpress inside to ensure consistency. Tasted great too.

And as for brew weight, i didn’t even know there was supposed to be a correlation, i only measure it for working out the equation :s

Hi Nico,

I tested with just one Kalita dripper. Interestingly, I just ran across a Gino filter (same size as Kalita) that had a flaw, a “thread bare” section in the bottom! We tend to just trust that the filters are all manufactured the same, but in reality, there are variances and when that happens, we blame everything else but the filter. Which is reasonable 99% of the time, except for the one percent where it isn’t. Doesn’t really matter day to day, but if there is a competition, definitely inspect filters at the very least.

I read through my original post and had a few “edit” issues which I corrected, which may help to be more clear. The super-automation was in response to coffee brewing, not espresso. While I agree with you on super-automation making an espresso that can target the same EDS and EY% much more easily than a human, there is also a large issue that the “perfect” coffee/espresso cup will never come from one of these practically speaking. A great cup, yes, the perfect cup, Never.

Absolutely never, in practical terms. Theoretically? Yes.

To achieve the perfect cup in a super-automated machine means incorporating an EK43 level grinder, which, will not happen.

Now to take the opposite position :grin:, philosophically :wink:, I make the supposition that given a super-automated machine with infinite adjustability across all variables that regardless of coffee grind quality, a “window” of recipe can be found to make the “perfect” cup. It is just the “window” for that cup may be incredibly small. But finite. Anyway, just a random thought, possibly that this can be proven mathematically…

Anyways, super-automation will not ever practically deliver the perfect cup, but can it deliver a great one? Sure, but will that be accessible to the masses? Likely no, just for practical considerations of cost and that the public just doesn’t care that much and can’t differentiate. Sort of like marketing Callaway golf clubs to me, given my level of play, pointless.

Regarding brew weight, another poster was referring to that as a means of insuring consistency. The reality is that, I believe almost any technique can be made to work, the only issue is how wide are the “windows” of optimal brewing, freedom to not be precise and yet get a great cup. It is why Cafes tend to always underextract to stay within an acceptable safety margin within that window, which is what seems to offend Rao :smiley: