Lose Weight, Gain Volume: About Coffee Bean Density in Artisan

Roasting coffee beans, you want to be as consistent as you can, replicating a successful roast while collecting enough data (besides what you see, feel, smell and taste from the coffee) to get an idea what you may be doing wrong if the results are not so good all of a sudden.

The Artisan software for coffee roasters has a feature to help you keep track of volume and density of your beans. Density is a word for the weight of a certain volume of beans and it can be measured in grams per liter.

Your green beans are obviously more dense than the roasted beans: roasted beans have grown a lot in size and they have lost weight (moisture) in the process.

In his new book “The Coffee Roaster’s Companion”, the famous Scott Rao explains these things: 

“Ideally, water should account for 10.5%-11.5% of green-coffee weight.” (p.3)
“Coffee loses 12-24% of its weight during roasting, depending on initial moisture content, roast degree and inner-bean development during roasting. The lightest palatable beans are probably those dropped during the latter stages of first crack and typically have weight loss, or shrinkage, of 11%-13%. About 30 seconds after first crack ends, shrinkage is roughly 14%-16% while at the onset of second crack, shrinkage is around 17%-18%. Dark, oily roasts may have shrinkage of 22% or more. The light roasts currently popular in the specialty industry lose an average of 14%-16% of their initial weight.” (p. 13-14)
“As beans lose weight during roasting, they also expand 150%-190% of their original volume. The simultaneous loss of weight and gain in volume equates to a density loss of almost half.” (p. 14)

This how you start with these measurements:

First find a container (preferably tall & slender, like a pipe, not a dish) that would hold a considerable part of the batch of green beans that you are going to roast. Fill this container to the brim with water and weigh the water. This way you also know the volume.

I took this rarely used Aeropress:

It fits 226.1g and because I had a bit of a bulge of water on the top, I assume it's 225ml

Then see (in this case on the Acaia coffee scale) how much weight of green beans fit in that same (dried) tube:

In this case it is 151g. I added just so many beans that a nylon puck resting on it was not resting on the beans but on the plastic rim of the tube.

I roasted 200g of green beans and got 172g of beans out.

After roasting, I measured what weight of beans would fit in the tube:

So that's 78g of beans (of roast color #99 measured on the Tonino). You can see the nylon puck on the top middle of the picture.

Then, I filled in the relevant data in the Artisan software and Artisan figured out that at this rate, my 200g of green beans must have been 298ml in volume and the resulting 172g or roasted beans have a volume of 496ml. I used 11g to do the Tonino measurement so I will have just a bit more than 160g in my bag of roasted beans.

These results are transported by Artisan to the page with other roast data and here you see that Artisan calculated the weight loss for me at 14% and the volume gain at practically 40%. Density of the roasted beans is 347grams per liter. Those are very useful reference points for me to remember:

(NOTE: the backend algorithm for calculating volume has been changed / improved in later Artisan versions)

Artisan displays this new intelligence (date, grams in, weight loss, Tonino value) below the graph:

And I print these on the new coffee bag label:

Then, if someone tells me they liked their bag of beans a lot, or they had some trouble with it (flow too fast for instance) they only need to tell me what's on their label and I can see what I have done  to this specific batch of beans. Same thing if someone tels me they found another bag especially nice.

And if I get a bag of beans that measures similar to beans I have already worked with, I can use the existing data as a starting point in looking for the sweet spot where these beans will be beautiful as well.

Thanks to Marko Luther for explaining this new feature to me!

PS Meanwhile I upgraded to a laboratory grade measuring cylinder:

PS 6 April 2015:

I also tried a different method to measure the density of green beans. Adriaan, a friend of mine, advised me to submerge the beans to get the closest approximation of the volume for a given weight of beans. I filled the glass tube with 100ml of water and added 100g of green beans. This added 92ml to the volume in the tube.

To get a good comparison between this value and the value of the volume of the roasted beans, I would have to submerge those as well but I feel that would be a waste. Also, the roasted beans would maybe absorb water quickly and thus change the measurement.