Hi everyone!

For several months I’ve been scratching my head to understand why the use of a paper filter below the coffee puck increases so much the extraction yield of espresso shots:

Scott Rao hit 25% extraction with his DE1

After reading that TDS measurements could be influenced by the extraction of sweet compounds from the paper itself, I stopped worrying too much about this topic.

A few weeks later, following Damian’s advice to use a convex shaped tamper in order to improve the consistency of the extractions, I ordered and tested several bases of The Force Tamper.

If the espresso extractions were indeed more uniform visually with the US curve base, I noticed a systematic loss of Extraction Yield (say about 1 unit).

In order to clarify the reasons for these variations, I imagined that it should be possible to further extract the spent pucks (with immersion brewing) after pulling the espresso shot.

I chose to divide my pucks into 3 concentric zones with cookie cutters: the corresponding surface areas / weight represent roughly 45% (center), 30% (middle) and 25% (outside) of the pucks

Each of the sections was extracted with the same technique (similar to a formal cupping) and the TDS of the resulting “brews” were measured.

With a standard puck prep and with a flat base tamper, a strong under-extraction of the outer part (periphery) of the puck has been measured.

The use of a paper filter under the puck considerably reduces this extraction deficit, which increases the average extraction yield:

A convex tamper base induces higher water flow in the center of the coffee bed and thus further reduces the extraction of the external grinds.

The resulting drink is a mixture of low extracted coffee from the outside of the puck and of from much highly extracted areas:

I will probably perform more comparative tests with other parameters or techniques (sieve baskets, blooming profiles, nutation…) and report if I find something of interest.

In addition, I think that this “study” could be a starting point to imagine new designs for filter baskets: for example I wonder if it is possible to expand the typical perforated area (round surface of 49 mm diameter) of the precision baskets and create a “high extraction” (Decent) filter basket.

Going from 49 mm to 53 mm would increase the perforated surface area by more than 15% and I suspect that the subsequent increase in extraction yield would be clearly noticeable!

More details in the pdf file:

Cheers, Stéphane