I am looking for advice on brew pressure and its possible effects on extraction. Here is my set-up: I have recently upgraded to an Andreja Premium and a Macap M4 stepless, bottomless portafilter, and including the family/friends deal, I pull 4-6 shots a day (I prefer ristrettos). Not being a pro-barista, I realize a lot of things can go wrong in tamping, grinder adjustment, etc. Sometimes I am able to pull some sound shots, sometimes not. The "not" part is my concern. Worst case scenario, I have noticed getting a couple to a few stalactites on the bottom of the PF that hang around (and not a consistent cone and gloppy drop formation; sometimes I get drops over the cup). I also notice sometimes little pinholes in the puck (incidentally?). Someone told me that I could be pulling the shot a too high of a brew pressure. When I backflush, the needle reads 10.5 or a little bit higher, and after I release it (end the backflush), it reads 10 exactly. I just pulled a shot to double check and it seemed to brew it at exactly 10.

My question is whether aside from variables in technique, grind, tamp, would I be better off with a slightly lower brew pressure? Perhaps it is just more and more practice!
Thanks,

Thanks, Dan, for the reply and advice. I will definitely lower the brew pressure and pull more shots. I remember reading the "Tweaking the Triple Ristretto" article, and while I have a triple basket just for such purposes, I probably need to work on "perfecting" (hmmm) a double basket extraction and varying the grind. I have also gained much from your "Perfecting the Naked Extraction."

I do have a related question, however; if this is a repeated question, perhaps you could point me in the right direction. What more generally is the effect of brew pressure on an extraction and what is it doing to the puck? I am relatively familiar with the differences in taste regarding temperature (say, pulling a shot at 204 and 199), but now I would like to understand more about brew pressure (and its possible relation to the properly (or improperly) tamped puck and brew temperatures).

HB wrote:I would lower the brew pressure to 8.5 to 9.0 bar....Though my experiments for pump espresso machines are all within the 7.5 to 9.5 bar range.

Is this the machine gauge or the gauge on a Scace? 10bar on the Vibiemme gauge is about 9bar on a blind basket and ~8.5bar on a Scace or an actual shot. 8.5bar on the machine gauge would give ~7bar on a shot.

[Edit : HB pressure measurements are with a Scace II which simulates puck back pressure.
Quoting Greg Scace, developer of the Scace Devices, from Digital Pressure Adapter :
"Yeah, the pressure difference between pumping against a blind filter and pumping thru a coffee cake is often very large. It's also machine specific, varying between types of machines, and sizes of gicleurs. I tested a machine that had a 4 bar difference between static pressure and dynamic pressure. Measurements were made using the same gauge, so gauge calibration wasn't an issue. "]

It really depends on the coffee - and on the extraction profile you're shooting for.

That being said - to use as a simple example a double shot of Hairbender updosed to 19grams, temp at 198F pulled a little short in around 27 seconds from a temp stable machine...

While pressure has definite impact on flavour profile, the easiest way to describe the difference is by focusing on texture. Higher pressure shots are "lighter" on the palate and seem to coat less, while lower pressure shots can have a "gritty" texture and are often very syrupy.

At a flavour level, with the above profile the shots tend to be more transparent but also less rich and dominated by the high end of the midrange flavours when pulled at a higher pressure. When pulled at a lower pressure the shots are "rounder" with less high end but better aromatics.

Let's see if I have it by way of summary. Please correct me if I am wrong. There seem to be two sets of issues here, (1) the one concerning the mechanics of measuring brew pressure, and (2) the closely related issue concerning the experience of what's happening with brew pressure on espresso.

1. Mechanical Concerns. Thanks for your replies and references Dan and Ian. Reading more of the threads you have highlighted, I understand there to be a difference in brew pressure between the brew pressure measured with a blind PF, which is "static," and the brew pressure that is "dynamic," that is, taken with the puck that itself can vary according to grind, amount, and tamp. We are concerned with dynamic pressure ultimately, and when setting the expansion valve, for example, we are really after the dynamic brew pressure even if we do it (or have to do it) according to a static read.

There may also be important differences between the Group Brew Pressure and the Machine Brew Pressure. The former, the Group Brew Pressure, can be measured by the Scace 2 device, and might be called (for lack of a better term) an "absolute" measurement because it delivers a reading right there on the group head. So now a question: In the case of an HX machine like my own, an Andreja Premium, will the Machine Brew Pressure (for lack of a better term) be more "relative" because it is gauged just before it goes into the heat exchange system? In any case, the Machine Brew Pressure might be called "relative" because the readings can vary from machine to machine (say, an HX and double boiler) and may not correspond to the reading from the Group Brew Pressure.

As I understand it, the actual numbers of brew pressure from the gauges are not important in themselves, but are significant insofar as they give us a consistent point of reference. Assuming that one's own machine is consistent—and speaking of one machine only—the reading of the brew pressure is arbitrary but functional so long as it can serve as a stable point of reference. But if we want to speak "across" machines, and as Ian remarks, help another diagnose a problem or give a "social" reference point for dialing in a shot with a specific blend, then the matter becomes more difficult. We would want a consistent way of measuring brew pressure across machines, and it seems that the Scace device could serve this purpose—assuming they are calibrated the same way (of course, we would all have to own one ).

So, I take from this a few things: First, when I am speaking with others about brew pressure, it is important to know how they are measuring it. Second, it may be possible that my 10 or 10.5 bar might be a "group" 9.0 or 9.5 bar brew pressure. Again, this only matters as a point of reference. Third, once we have this straight, it can be meaningful to speak of optimal dynamic brew pressures for such and such a blend, or perhaps to help diagnose an extraction problem as a "brew pressure" problem. Does this sound right?

2. Experiential Concerns. I realize that none of this can replace an experienced barista and a trained palate—it's all about the taste!--and that's the point of the brew pressure measurements. Chris's comments about texture and flavor were really helpful.

I am familiar with Hairbender (originally from Portland), and when I was in Portland last month on my migratory trip to the Division St. Stumptown, I had a couple ristrettos. I sensed (in my newbie way) what was probably more on the higher brew pressure side (though I realize there are a lot of variables)—actually it seemed a little nuttier (almost almond-like). In any case, I have been fixed on B and B's Dancing Goats, and have a couple more lbs coming in shortly. I will try to experiment with brew pressure with the newly arrived beans. Do you have any suggestions concerning what I should be attentive to with this blend?
I am eager to learn as an amateur more about this variable, brew pressure, general telltale signs of high or low brew pressure, and from those more experienced than I, what brew pressures do with different origins and blends.

Anthony -

Some more light reading:

experiments-in-programmable-variable-brew-pressure-profiling-t3478.html
brew-pressure-profiling-update-2-t3566.html
brew-pressure-profiling-update-3-t3703.html
playing-with-pump-pressure-part-uno-t4812.html
playing-with-pump-pressure-parte-due-t4888.html

Your Andreja measures brew pressure at the hx outlet and the difference between what your gage reads and the pressure at the puck is 0.3 to 0.4 bar. Adjusting your OPV such that the gage reads 9.0 with a blind filter would work well. Most home users (me included) adjust their OPV as such and rarely change same. The OPV serves to regulate the maximum pressure on the gage. However, from your well written posts, I sense you want to play

Trying to discern the difference that a variation in brew pressure causes is an arduous task. For a specific bean, I would suggest you obtain a 0.1 gram scale and use the same amount of grinds for each pressure variation and extract the same amount of espresso for each trial. If you adjust the OPV in, say, 0.5 bar increments, you should find that you will be making grinder adjustments to keep the extraction volume constant (for the same extraction time). It gets complicated but keep us posted on your findings.

Anthony wrote:I am eager to learn as an amateur more about this variable, brew pressure, general telltale signs of high or low brew pressure, and from those more experienced than I, what brew pressures do with different origins and blends.

Over the course of the first year I owned the Vibiemme, I frequently adjusted the brew pressure to learn the effect. By that time, it was pretty obvious that ±0.5bar is insignificant relative to freshness, dose and temperature. Eric's thermocouple adapter prompted me to cobble up the Digital Pressure Gauge, but mainly for academic interest to see where I'd settled by taste over the course of the previous year's fiddling. Other than resetting it after a lube job, I haven't changed the OPV setting in more than a year.

Today I mainly use the DPG to set / reset a machine in the 8.5-9.0 bar range on the puck, regardless of machine gauge reading.

Andueza and Maeztu, at the food science department at the University of Navarra, did a series of studies on espresso taste, including one that taste tested at various pressures. There wasn't a huge difference, but 11 bar scored lower than 9 bar or 7 bar due to added bitterness, while the 9 and 7 bar shots were almost too close to call. The rough grain of these results make people's claims of big differences for 1/4 bar pressure adjustments somewhat suspect.

erics wrote:Anthony -

Some more light reading:



Your Andreja measures brew pressure at the hx outlet and the difference between what your gage reads and the pressure at the puck is 0.3 to 0.4 bar. Adjusting your OPV such that the gage reads 9.0 with a blind filter would work well. Most home users (me included) adjust their OPV as such and rarely change same. The OPV serves to regulate the maximum pressure on the gage. However, from your well written posts, I sense you want to play

Trying to discern the difference that a variation in brew pressure causes is an arduous task. For a specific bean, I would suggest you obtain a 0.1 gram scale and use the same amount of grinds for each pressure variation and extract the same amount of espresso for each trial. If you adjust the OPV in, say, 0.5 bar increments, you should find that you will be making grinder adjustments to keep the extraction volume constant (for the same extraction time). It gets complicated but keep us posted on your findings.

Thanks for the references, Eric, the info, and the encouragement!

Anthony wrote:I am familiar with Hairbender (originally from Portland), and when I was in Portland last month on my migratory trip to the Division St. Stumptown, I had a couple ristrettos. I sensed (in my newbie way) what was probably more on the higher brew pressure side (though I realize there are a lot of variables)—actually it seemed a litter nuttier (almost almond-like).

Division Stumptown machine (used to be) run at 8.75BAR

Oh, I see. So, to understand better what you were writing earlier, would you say this is something like a middle-high pressure? In any case, I assume this setting on that machine works well for Hairbender, or what the Baristas there want to bring out in Hairbender.

Thanks for the references you posted, Jim. More edifying nighttime/weekend reading!

To start off my little quest, I have just lowered my brew pressure to a static 9.5 bar—beginning there. I am overdosing and pulling normals and ristrettos (Dancing Goats about 5-6 days from roast date, at about 200 F--so pulling the shot with the digital read at the adapter site at about 203-204 F). For the sake of consistency and to see what would happen, I pulled a couple shots with the old brew pressure with these beans (at 10.5 bar); I did not adjust the grind at first with the new brew pressure, but it feels like I want to tighten it a bit. (This seems counter intuitive, but there it is.) I am not having the "stalactite" problem so far at all, and I do not see any pinholes like I was seeing before [I should say, it was not a problem with these beans even at the higher bar this time, but was a problem a few days ago.].

I just don't have the refined taste buds and vocabulary to express the experiences, but something changed. For example, the texture was a little denser and had a softer finish, not elongated, and was more "forward" on my palate than before (even a little "fruiter"). There was no astringent after taste like I was getting sometimes previously. And at this brew pressure, the tiger striping even seems to be a little deeper and I am not getting what seemed to me before to be a pre-mature blonding with a normal shot.

Let me ask a question that seems so evident as to be naive, but it is a matter of confirmation. When I am pulling a ristretto, my brew pressure is right around the static setting (like I would have with a blind PF), and when I am pulling a normal, it is around 9.0 bar, so 0.5 bar less. Does that sound right?

Not trying to set my taste-buds according to brew pressure, but eventually my brew pressure according to taste, for now, nevertheless, I am wondering "what happens when ...".

Anthony wrote: Let me ask a question that seems so evident as to be naive, but it is a matter of confirmation. When I am pulling a ristretto, my brew pressure is right around the static setting (like I would have with a blind PF), and when I am pulling a normal, it is around 9.0 bar, so 0.5 bar less. Does that sound right?.

Yep. That sounds right. There are two things at play here. First, when no water is flowing through a plumbing system, the pressure is the same throughout. In order to get water to flow through a pipe you have to have a pressure differential across the water. Imagine that you are holding your finger over a garden hose so that no water squirts out. The force of your finger pressed on the end of the hose exactly opposes the water mains pressure and the net force on the water is zero, meaning that no water moves down the hose. Once you release your finger, the pressure at the open end of the hose is at atmospheric pressure, while the pressure at the water mains connection is the water mains pressure. Now there's s big force imbalance and water flies out of the pipe. In the case of your espresso machine, the flow when pouring a ristretto is pretty small, so the measured pressure is gonna be similar to the static case (when pumping against a blind filter). As soon as substantial water flow occurs, such as when brewing normal espresso, the pressure drop along the flow path increases, and the pressure at the group is reduced. The magnitude of the pressure drop that is reported by the gauge in your machine depends on where in the flow path the gauge is tied into the system. If it's close to the OPV you may not see much difference (but read the next section). If it's near to the group you'll for sure see a substantial difference.

Second, we have to consider the behaviour of pressure relief valves that rely on spring pressure and a plunger to regulate the pressure of a system. Spring force on the plunger is given as k*X, where k is the spring constant, given in pounds per inch of displacement (or Newtons per meter in SI units), and X is the amount that the spring is compressed from its free, uncompressed length. The important concept is that the more you compress the spring, the more force the spring exerts. When you are pumping on a blind filter, or pulling a ristretto, the spring is compressed more than when you are pulling a normal shot, because more water is being forced through the OPV. As a result the spring puts more pressure on the plunger and the OPV regulates at a higher pressure than when less water is flowing through the OPV, the case when relatively more water flows through the group. Depending on the OPV design,and the capacity of the pump, the difference in regulated pressure can be pretty large. I once measured a system of vibe pump and OPV that produce 4 bars difference in regulated pressure between pumping on a blind filter and actually brewing coffee.

For espresso machines, the flow-dependent pressure drop through the plumbing system and the difference in regulated pressure due to relative changes in spring force imparted to the OPV plunger work in the same direction, unfortunately toward more pressure variability. As relatively more of the total water flow goes to the group, the regulated pressure (from the OPV) goes down, and the pressure is further reduced as the water flows down the flow path to the coffee. The effect is greater with vibe pump machines than with rotary pump machines. In rotary pump machines, the internal pressure relief valve in the pump has a very large bypass port and the pump itself has a very high flow capacity compared to a vibe pump. The difference in flow between a ristretto and a normal espresso does not change the position of the plunger much, resulting in less difference in regulated pressure compared to vibe pump / OPV systems.

Constant pressure regulation requires alternative regulation techniques, such as a system of variable speed pump, electronic pressure transducer, and process controller. One of the advantages of the system that Eric S referred to in his light reading suggestions is that the regulated pressure at the pump is highly reproducible, resulting in greater ease in maintaining consistency.

Getting to the real nut of the matter - The idiosyncrasies of espresso machines are interesting if you are so inclined, or if you are a machine designer or engineer. It's often useful to know why things work the way that they do and to be aware of how and why variability is introduced into the brewing process. I don't know what you're capable of mechanically or electronically, or if you're a super-taster or what, but at some point you'll prolly wanna decide for yourself what you can and can't taste, how much of an effect pressure variability is on your particular setup, and how much money and effort you wanna spend on seeing if you can do better. Generally pressure variability is not the tallest nail sticking out of the board. Whack on the tall ones first, taste carefully, draw your own conclusions. You seem pretty smart.

-Greg

Greg:

Thanks for taking the time to respond and to explain these points on brew pressure, really, with such clarity. Would constant pressure regulation eventually have to take into account the variable resistance of the puck after it is infused, or rather, during the process of infusion?

To your final comments: Of course, eventually I would like to produce (and reproduce) great tasting coffee, but getting there is also pretty fun, too! I have already found brew pressure significant enough to alter the taste at least within one bar, and this is something I never really knew before, though your reminder that brew pressure is not the most significant variable is helpful. What I find intriguing about brew pressure is that it relates to the environment in which espresso is produced. As such, it appears to be one of the few things—after one figures out what it is doing, perhaps through some experimentation—that can remain more or less part of a stable environment when dialing in other things.

Aside from this, the nuances of espresso are fascinating, and I admire the ability of those who can find their way around the subtleties of coffee either intuitively/experientially and/or systematically. Finally, to invoke Socrates (though outside of Hemlock, I am not sure what he drank!), the unreflective life is not worth living; in that spirit, pulling a good shot entails for me in part also knowing, as you put it, "why things work the way that they do and to be aware of how and why variability is introduced into the brewing process." (Though sometimes, heck, I'll just settle for tasting a good shot.)

Anthony

After some experimentation, I have followed the suggestion of several here (again, thanks for the advice) and lowered my brew pressure again, now to a static 9.0 bar (according to the on board gauge). I have been noticing a consistently more viscous ristretto (sometimes within a more nutty/carmelly range) across several different blends with which I am familiar. I have also noticed far fewer problems with channeling compared to when it was set at 10.5 bar.

This brings me to my simple question. Why do the distributors of such espresso machines pre-set the brew pressure for something like 10.5 bar and not e.g., 9.0 or 9.5 bar? They cannot be assuming that we are all making lovely triple basket ristrettos like Jon Rosenthal!

Anthony

The pressure is set Hi to avoid "My machine doesn't work" tech calls.
At 11 bar, with commercial coffee or sub any grinder we'd be caught dead with, you'll always get something in the cup.

It's kind of like buying a new target pistol. Those that know what they're doing will sight it in.
Those that don't can't hit the broad side of a barn unless they're standing on the latch.

cafeIKE wrote:The pressure is set Hi to avoid "My machine doesn't work" tech calls.
At 11 bar, with commercial coffee or sub any grinder we'd be caught dead with, you'll always get something in the cup.

And you would know this ... how?

Here's the real story, once again.

• The desirable pump pressure for espresso is roughly 8 to 9 bar. Below that, one loses crema, above that one gets extra bitterness.
• Espresso machines are routinely adjusted against a blind filter. This leads to the two problems Greg mentions: first the pump puts out less pressure at espresso flow than at zero flow, and second, there is flow resistance between the pump and the puck, which can drop the pressure before the puck is reached.
  Vibration pumps have a steeply declining pump curve that puts them roughly in the right range for doubles, but not for singles. Rotary pumps have very flat pump curves
• Because of its flat curve, a rotary pump is adjusted to around 9 bar against a blind filter. Flow resistance will drop this pressure, but not by much, and it will remain between 8 and 9 bar.
• Vibe pumps are adjusted to 11 bar, since the pump curve will drop them by 2 bar in any case.
• OPVs, the valve which allows the pressure to be adjusted, are variable in quality, especially the ones for vibe pumps, as Greg found out. Adjusting the pressure at the high end compensates for possible defects.

The service and adjustment standards for espresso machines, as done in Italy, may not be perfect, but they are based on long experience. Only a fool would assume they are systematically wrong. Sometimes, Italian manufacturers make odd assumptions about the US market, and deliver machines not tuned as well as they would be in Europe. This is not the case for the 9 bar and 11 bar settings for rotary and vibe pumps, which is the standard in Europe as well.

another_jim wrote:And you would know this ... how?

Horse's mouth. From more than one manufacturer.