I wanted to know how much of what's coffee is extracted from the beans. There is lots written on this, I didn't feel it gave sufficient details and so I decided to have my own go at this. I thought I'd start with the same thing and go two different ways.
Methodologies
Method one was to take the puck, weigh it first (in the basket/groupset tamped), extract the coffee and then dry it and weigh it after.
Method two was to take the coffee I'd made, then remove the water (drying it) and weigh the remainder.
I did both.
Baseline:
Method 1
I weighed the basket (the group head steel basket), then added my usual amount of coffee (my grinder has a memory and produces pretty similar deliveries of grinds), then weighed the combined (tamped) basket and coffee, subtracted the basket and had 17.93g of coffee grinds. (side note: I also use this to work out my ratio for making a flat white as I do)
I banged out the puck (biscuit) and put it in the oven in a plastic bowl (11.7g) for drying at about 47C.
I kept inspecting and weighing and breaking the puck up and stirring with a fork, until eventually it became nice and dry
So ultimately I seemed to lose 1.8g from the puck. Interestingly this result gave about 1g of loss per 10g of coffee, as identified in the first AI answer.
Concerns
I was thinking while doing this (the iterative stirring, checking and replacing) that I had no idea when was "Dry" and what the actual state of the water content of the beans was before I ground them. Because I knew that would be about 3% and how would I know what the water content of my drying was?
This was vexing and so I decided to approach it from the other end: what could be found by removing the water from the espresso.
Method 2
This method was similar but less measurement intensive. I would pull a shot, into the above demitasse and then weight would tell me how much coffee I had to start with (57.3g) (from 17.45g of beans) which is a bit more than I pull into my mug (I don't weigh, I use volumetric on the mug and judge that by taste then use the time for a pull to make it a bit more consistent)
I soon saw that it would take too long to dry that coffee out, so I tipped it into a shallow small dish that would allow faster drying (because surface area to volume ratio) and put both the cup (with some tiny amount of coffee in it) and the saucer into the oven for drying. I dried till it looked like freshly applied paint (but didn't touch it); not moving around at all (as liquids do) when inspected.
this gave a different result 4.68g (interestingly similar to the second AI answer). However what I don't know is what amount of "fines" came through (to add to that mass of coffee) and how much of water may remain bound to those coffee remains (its very hard to remove the water from a solution).
Conclusions
So now I have two answers of my own making and in some ways have not got just one answer. Reflection on this has led me to wonder if the best approach is to repeat method 1, but add in a "control" where I have ground coffee of the same mass and only put one through the espresso extraction process ,but both through the oven. This would mean that I could track the weight of the "control" grinds against the espressed grinds. Any loss of weight of the control sample could then be further subtracted from the final weight of the espressed grinds and I'd have an answer.
The only question at hand now is DIGAF
