Pump pressure is an extraction variable that is discussed from time to time. Many people make the reasonable assumption that a higher pump pressure will result in a faster flow rate. Espresso is anything but reasonable, however. In Illy's book, Petracco has a graph (p. 160 of first edition, p. 267 of second edition) that shows a faster flow rate at 5 bar than at either 3 bar or 7 bar.

Since we generally don't extract espresso at 3-7 bar, I decided to do a little testing of my own. Fortunately I own two devices that made the project a little easier, the Schectermatic profiling pump and the Schectermatic Evaluspromatic (shoutout to Mark P!)

I pulled a series of shots at 201F using Coffee Klatch WBC blend, dosed at 13.5 - 13.6 grams. Grinder was the 3-phase Robur running at 40 hz. Brewing ratios were in the 60-70% range.

The profiling pump was set to preinfuse at ~3 bar for 4 secs, then rise fairly rapidly to either 8.0 bar, 8.6 bar, or 9.1 bar. With the Evaluspromatic I divided the shots into six segments and weighed the amount of espresso produced in each time period (the first segment was uniformly zero).

The data show that an espresso cake is anything but a simple flow restrictor. It appears that fines migration and/or compaction due to pump pressure makes it a very complex restrictor indeed. In fact, the meager amount of data that I collected suggests that in the normal espresso range, higher pump pressures produce slightly lower flow rates. This is more-or-less consistent with Petracco's data, if one cares to extrapolate from it.

Some people obsess about the effect extraction pressure has on flavor. There may be some significant effects to be found there, especially by manipulating the pressure profile in the crucial early part of the extraction. But I haven't gotten very far into that yet, and I think it will take a blind taste panel to prove that significant changes can be wrought.

Oh yeah, shoutouts to Greg S, John E, and Sean L, who have all built their own profiling pumps. Greg, in particular, has posted a bunch on this topic here on home-barista.com.

I've been a big advocate of taste testing when doing this type of experiment. But looking at this (and after all the grinder testing), I'm beginning to wonder.

It looks like for this setup, the 9.1 bar shots produced the least variance in the pair of shots, although it'll take more tests to see if the pattern holds. I'm beginning to think that consistency rather than taste may be what determines the reputation of espresso equipment. The standard variable settings for espresso shots may be where that, rather than any taste factor, is maximized.

In terms of taste, my micro-economics trained instinct would be to look at the maxima, minima, and inflection points. If the fastest flow is around 6 bar, perhaps a series of shots using finer grind to yield roughly equal flow rates at that pressure would show some aspect of the taste getting maximized or minimized at that point. Of course, it may not be a desirable aspect of the taste getting maximized or an undesirable one getting minimized

Whenever I think of Andy Schecter, I see this giant brain, about to explode out of its cranial vault, but that speaks not

It is refreshing to see Andy not merely doing some research, but actually posting the results here for us less well endowed folks to see!

Keep it up; I'm sure you will discover something very important, and we are all reading anxiously with receptive minds!

ken

Andy:

I'm glad you're doing it and not me. After Brew pressure profiling update 3, I sorta liked what i was doing, decided that convincing the rest of the world was gonna be the biggest headache ever, and just sat back enjoying the results. Lemme know if I can help a bit. I'm pretty booked up until the end of October, but after that I can revisit and duplicate results on different equipment if you so desire.

Right now I'm training for the Northshore Marathon (inline speedskating race in Duluth with 4000 skaters), and getting ready for the next WBC machine selection trials.

Gotta go fast! I'm very interested in what you come up with.

-Greg

another_jim wrote:It looks like for this setup, the 9.1 bar shots produced the least variance in the pair of shots, although it'll take more tests to see if the pattern holds. I'm beginning to think that consistency rather than taste may be what determines the reputation of espresso equipment. The standard variable settings for espresso shots may be where that, rather than any taste factor, is maximized.

In terms of taste, my micro-economics trained instinct would be to look at the maxima, minima, and inflection points. If the fastest flow is around 6 bar, perhaps a series of shots using finer grind to yield roughly equal flow rates at that pressure would show some aspect of the taste getting maximized or minimized at that point. Of course, it may not be a desirable aspect of the taste getting maximized or an undesirable one getting minimized

Jim, thanks for your interesting comments.
1. Yes, the seeming linear relationship between pressure and shot variance is intriguing, but will it hold up to further trials?
2. Consistency rules; everyone who goes from Silvia to E61 box seems to comment on more consistent shots.
3. I should continue the series at lower and lower pressures to see where the maximum flow rate really lies. Probably won't have the best crema, though, at that point.

The feeling I have about this pressure stuff is that there's SO MUCH we don't know about how pressure and the puck react. Until I understand the whole thing a little better, I don't really feel that motivated to do a lot of formal taste testing. Too many blind alleys right now....

gscace wrote:I'm pretty booked up until the end of October, but after that I can revisit and duplicate results on different equipment if you so desire.

Very cool, Greg, that would be great.

Ken Fox wrote: I see this giant brain, about to explode out of its cranial vault, but that speaks not

Like the old Outer Limits episode where they rescued some guy's brain and one eyeball, then hooked them up to a nutrient bath that kept them functioning?

AndyS wrote:Very cool, Greg, that would be great.

One thing I can tell you - It's been my experience that higher pressure, and fast pressure rampup both contribute to slower flow.

-Greg

gscace wrote:One thing I can tell you - It's been my experience that higher pressure, and fast pressure rampup both contribute to slower flow.

-Greg

Hi Greg,

Is this PF dose dependent? In other words, do you get the same observation with a standard, Italian-ish dose of 12 or 14g, as you would get with 18 or 20?

ken

Ken Fox wrote:Hi Greg,

Is this PF dose dependent? In other words, do you get the same observation with a standard, Italian-ish dose of 12 or 14g, as you would get with 18 or 20?

ken

Good question. I presume so, but I don't really know. I can find out easily enough next month, when I get a little time.