MARGINAL GAINS - MAXIMISING EXTRACTION THROUGH FLOW & PRESSURE

No area of coffee making lends itself quite as well to the theory of Marginal Gains as extraction yield. An inordinate amount of time from some of the most innovative thinkers in coffee is spent pushing the limits of techniques and equipment, all in the name of that extra few percent of extraction. The Why of this is well docmented, most notably by Matt Perger on his Barista Hustle site  http://baristahustle.com/. Its in the How that we keep exploring new avenues for improvement.

The last month has seen us exploring how brew pressure and flow rate can help us maximise our extraction potential, its not something I can claim to have thought of as a concept, that goes back to a conversation had by Maxwell Colonna-Dashwood, but we rehashed those thoughts after the fact and both implemented them to startling effect.

Now is probably a good time to catch up on the origins of this line of thought by reading these two posts;

https://colonnaandsmalls.wordpress.com/2015/05/10/15g-baskets/

https://colonnaandsmalls.wordpress.com/2015/05/26/pressure-flow-restriction/

Maxwell ably describes how changing the pressure, and so the effective flow rate, dramatically influences his extraction potential. In the second post he succinctly breaks down how pressure and flow rate are interlinked, this is an application of Bernoulli's principle of fluid flow where pressure and velocity of flow change with the cross sectional area of the pipe.

How does this help us eek out that little extra extraction? I believe the answer stems from how the pump pressure/flowrate we choose affects the nature of the flow dynamics and so the brewing within the puck. 

Espresso has traditionally been brewed at high pressure provided by either a spring and piston system or, as is more common, a rotary or vibratory pump. For pump machines the standard system pressure for brewing has been for years and still is 9 bar. Scott Rao advocates tuning the pump pressure to deliver the maximum dose by weight over a set time, the theory being that beyond this point we see some compression of the puck and so increased resistance. This approach should deliver a pump pressure for optimal dose weight for a given system, but any change in the system (coffee, dry dose, grind etc) will result in a change to the optimal pump pressure. It does not follow though that this will result in maximum or optimal extraction. That's because extraction is effected by so many more variables than just your pump pressure.

At this point we could spend a lot of time going over the known theories on how the puck reacts to water flow, but I'd like to keep this related to our more recent experiences.

If you've read any of my posts on using the Strada, you will know that we run a constantly changing pressure profile that peaks at 15secs then ramps slowly down from there. Over the past year we've seen our peak pressure move down from 12bar -> 8bar -> 6bar. The move from 12 to 8 bar was originally taste driven, we found the acidity easier to balance at this lower pressure. The move to 6 bar was driven by the conversation I referenced above and it's made a marked difference to our extraction potential.

For the first 4 months of 2015, our average extraction has been 20.85%. Since the beginning of May and the switch to a 6bar peak profile, our average extraction has been 21.28%. In fact, extractions below 21% are now in the minority rather than the normal occurrence.

So what do we think is happening here? 

It helps if we consider some of the basic factors that will affect fluid flow through what is a granular bed. What we are interested in here is the flow rate of the water through the bed, and we can think of this as the velocity a volume of water has passing a certain point. Seeing that we know the cross sectional area that the water flows through, we can consider the velocity as our indication of flow rate.  

It has been shown that the velocity is directly proportional to the pressure drop across the bed and inversely proportional to the bed depth, this is known as Darcy's Law. What this means is that if we reduce the pressure drop or increase the bed depth, we also reduce the flow rate. This leads me to believe is that for each system and each dry dose (and so bed depth) there will be a corresponding flowrate for optimal extraction, which we set primarily via our pump pressure. In our system, Strada/Mythos/Pergtamp with light roasted, well developed single origins the lower flow rate peaking at 6bar allows us to achieve extractions closer to this optimum.  I believe this stems from a more even extraction across the bed, coupled with already high extraction potential.

I have several ideas as to why this is;

• The slower flow rate gives us a longer dwell time, thereby allowing for a more efficient brewing process and each particle achieving a higher extraction. This is analogous to the later stages of filter brewing, where time and temperature are crucial to achieving efficient extraction

• The slower flow rate and pressure could facilitate more effective off-gassing, this would explain a richer crema rather than the reduced crema one might expect. Simply put, if we are extracting more fully, then we are allowing more of the CO2 to escape allowing the water access to more of the coffee.

• I see puck density as a zero sum equation, for every area of low density, you will have areas of high density to compensate. A more even puck expansion is one of the known benefits of pre-infusiuon. However, if we consider that our water flow will always tend to the easiest path, water driven by a lower pressure may not experience the same urgency to flow towards low density areas, ostensibly allowing better extraction from the higher density areas of the puck. Think of it as achieving an even extraction gradient throughout the puck.

• A slower flow rate should increase temperature stability. This in turn should lead to higher overall brew temperatures and so more efficient extraction.

From this it should be obvious that taken to its furthest extent, each individual set up of brewing system and coffee will have a unique flow rate and pump pressure for optimal extraction. Think about the complexity of coffee density changes altering bed depths, just this variable alone would be enough to make minute changes to the fluid flow, and thats before we consider the change in bed depth etc.... 

So where does this leave us?

Finding the optimal pressure/flow rate for every coffee you run is impractical in a shop setting, so the goal is to find that which provides a positive result most often. For instance, I know of three sites that have moved to lower pump pressures, all have seen a rise in average extraction %'s and all run different set ups. Interestingly all three are running different pump pressures.

The standard of 9 bar wasn't arrived at by accident, it worked for the style of roasting that has been prevalent for decades and is now "traditional". Surely it's logical that if we change how we roast, we should look again at all aspects of how we brew?