What is the optimum brew pressure for espresso?

Need help with equipment usage or want to share your latest discovery?
User avatar
Coffee Physics

#1: Post by Coffee Physics »

Can anyone give me advice/experience on what you consider is the optimum brew pressure for espresso extractions? I've read that 9-10 bar is the best but I haven't seen any descriptions of how the character of the shot changes as the pressure is dropped from say 12 bar down to less than 8 bar. I'm wondering if this depends on the water temperature, the kind of coffee, or the length of the shot. Do some coffees extract better at higher pressure while others perform better at lower pressures?


#2: Post by coffeedom »

You'll find a few discussions of similar questions related to pressures and pressure profiling on this site, though your question is more to the point.

This is a hard question to answer given the current knowledge of profiling. As you may know Illy's research led to 9 bar extractions being more or less standard, at least as far as pump machines go.

In order to get a full answer to this question, pressure profiling needs to be explored in depth and more scientifically. It's been happening, albeit slowly. There aren't a lot of machines that can do it unless you are willing to spend a lot of money, so most people are restricted to 9 bar profiles, 9 bar + preinfusion, or spring lever profiles. From these, many will say the spring lever declining profile produces the nicest tasting shot, in general. But that also depends on coffee, etc.

My personal feeling is that the 9 bar standard is probably best suited to lower doses, mainly singles. As dose increases, there is more room for extracting at lower pressures. I fortunately can profile each shot, and have found, as a general trend, that for higher doses exceeding 16g for a 25+g shot, I can really bring down the pressure and still get very syrupy and rich tasting shots, with a little less bitterness.

Some spring levers extract at 6 bar and decline a bit during the shot; they are also known to make great tasting shots. So there is room for lower pressures I think. Going above 9 or 10 bar is probably not useful unless you are extracting very small doses with very fine grinds. Nespresso does this with their 5g capsules. They need the higher pressure to simulate crema through a sort of venturi-like effect during extractions.

But for home baristas with real machines and coffee, I don't know if going over 9 bar has a lot of benefit. It seems to me the space between 5-9 bar is the optimal space, and then it depends on dose, coffee, degree of roast, temp, and ultimately the taste of the user. Over the next few years I think this question will become a bit better understood as more testing and experience accumulates.


#3: Post by MerleApAmber »

From your mouth to God's ear.
(or some grad student's who's willing to learn all there is about extraction and more - willing to share it, not bury it behind some corporate confidentiality agreement.)
Thanks for your thoughts.


#4: Post by coffeedom »

Yes, it's probably a bit of wishful thinking on my part! Profiling has been happening for a while now, with not much of a concensus about it. The technology is there, so we'll see what happens. There are more than a few members of this forum who may have a greater knowledge of this than I do and may offer a different perspective.

Extracting at around 9 bar is pretty much the norm and some do feel that, aside from preinfusion, playing with pressure doesn't really do much. That may end up as the prevailing view, who knows. My own feeling is that varying the extraction pressure (like we get with levers, for example) can make a difference but how much isn't clear.

User avatar
Coffee Physics (original poster)

#5: Post by Coffee Physics (original poster) »

Thank you for your very helpful reply. Could you elaborate a bit more on your approach to profiling - what machine do you use (what others might you recommend) and what is your strategy when pulling a shot? Can you actually monitor the pressure real time during the extraction?

User avatar
Coffee Physics (original poster)

#6: Post by Coffee Physics (original poster) »

I attended a La Marzocco "Pressure and Flow" workshop yesterday in San Francisco where they were showing their latest version of the La Strada EP, capable of pressure profiling shots tailored to specific SO coffees and roasts. We spent time pulling shots and experimenting with 4 different profiles, three of which were essentially triangular profiles, zero to peak (12bar, 9bar, and 6bar) and then back down to zero over a 30 sec period and the fourth which was just a steady 9 bar extraction. The theory is that the bean density determines the profile approach, more dense coffees (higher altitude grown and or lower roast temperature) working better at the higher extraction pressures while the less dense coffees (lower altitude grown and or higher roast temperature) do better at lower pressure.

The machines were impressive, capable of providing any custom pressure profile you would want and will soon be programmable with your laptop through a USB connection. I was amazed at how delicious the low pressure shots were - very sweet and syrupy, noticeably more to my particular preferred taste than the higher pressure extractions. The temperature was kept constant, around 200 degrees, for all the shots. While I don't think I'll ever have a Strada EP (!), I am going into my expobar to adjust the OPV down to see if I can improve my shots.


#7: Post by coffeedom »

Hi Richard,

Rather long response, apologies. The LM Strada is obviously one of the machines that can profile. Good that you could see it firsthand. Other top end machines capable of adjusting pressure, as you may already know, are the Slayer and Synesso machines. There's a new custom 2-group Slayer in Costa Mesa, if you are ever in SoCal. Looks awesome.

These machines use various means of achieving control, such as flow restrictors or direct control, each with pros and cons.

Other machines of interest include the Bezzera Strega, which allows for deviations from the spring lever profile, and the Breville Dual Boiler, which apparently allows pressure adjustment on the fly by playing with the hot water valve. There is a thread about this here on HB, but can't comment on it beyond that as I haven't tried it.

You can also try to put a 160 or 200 PSI pressure gauge on a manual lever to see what you are doing, though the level of control in either direction may not be great.

At home I have two machines I use for profiling: the Rossa HC Hand Espresso and the Rossa PG Air Espresso. Both are similar but have different approaches. Both allow you to adjust the pressure directly in real time. You can do pretty much whatever you want on both. The Air Espresso is pneumatically driven by pressurized air and the air pressure can be increased or decreased at will.

For single shot brewing, it's my favorite device. I absolutely love this thing. You can also do rough temperature profiling on both Rossa devices. It's easy enough to pull shots at any temp you like between say 88-96C (190-205F) and you can build in a temp decline if you like, or keep the temp steady during extractions.

This approach is fully manual - no USB devices or electronics. Just solid brass, pressure gauges, and the ability to measure temp directly (before extraction). The pressure is measured just above the coffee puck, so it's about as accurate as you can get.

My approach depends on the coffee. I'm intrigued by what you were told at the LM workshop about higher grown, harder beans needing more pressure, with lower grown or darker roasts working best at reduced or declining pressures. My experience hasn't led me to this conclusion but it could very well be true. I suppose the theory is that harder beans need more pressure to get the goodness out of them. Or something like that.

For me I'm finding a connection between dose and pressure. As dose increases (and grind coarsens), I find that I can bring the pressure down to help reduce harshness, something that can happen with coarser grinds. For a 20g dose and 35g extraction, something along the lines of a double ristretto, I slowly ramp up to 6 bar and hold it there for the duration. These shots are big, goopy, and surprisingly soft, though not usually as soft as a lower dosed shot with a finer grind. But the flavors are huge.

If I've roasted a Brazilian and maybe a Central American as well, then I'll go for a more traditional shot, with maybe 12g in a single basket and a quick ramp up to 9 bar for the duration. Of for single origins I'll maybe do a lever profile - brief pre-infusion at 2 bar, quick ramp up to 8.5 bar and a slow decline to 5 bar, ending at about 30 seconds.

Anyway the possibilities are endless with these devices, but they are not for everyone. I like that they are fully manual, with compete control over every aspect of extraction, including to some extent temperature. But you just can't walk up and pull shot after shot. I use the Rossa devices when I make espresso for myself, maybe 1-2 times a day. When I'm making coffee or milk based drinks for friends or family, I use other machines.

User avatar
Team HB

#8: Post by another_jim »

I've been using the Bezzera Strega, a pump/lever hyprid for about two years now; and have become convinced that declining pressure throughout the shot, after the flow starts, improves shot quality. If one uses a pump, the flow accelerates and gushes by the end of the shot. If you can lower the pressure smoothly, the flow remains even. Doing this manually rather than using a preset program allows for perfectly even flow; but even a preprogrammed declining profile will get close to maintaining even flow.

The effect on taste is simple. Espresso blend traditionally have muted acidity and roast flavors, so the coffees taste somewhat dull when brewed normally. This is because in espresso these high and ow flavors are amplified and distorted. This effect disapears when you drop the brew pressure and maintain even flow. The upshot is that you can use a much wider range of coffees and roasts for making shots. If SO espresso becomes more common, you will see most high end cafes that use them going to profiling machines.

There is a caveat here. Profiling to keep the flow steady extends the length of the shot from about 25 seconds to about 35 to 40 seconds. In most HX machines, the water would be overheated by the end of so long a shot; so this requires either a double boiler or classic lever machine (or a specialized group that holds the two ounces of water used for the shot).
Jim Schulman


#9: Post by Anvan »

While not claiming anything near Jim's background, palate or expertise, I can second his comments from my own experiences. Eight months ago I installed the manometer upgrade on the GS/3, and while the mod is a little sensitive and tricky to use, it does enable one to maintain pressure control throughout the shot.

While I've found some - at least I think so - improvements in the texture and richness of traditionally styled espresso blends (and especially Brazil-based blends roasted into full city or maybe beyond), the real surprise for me was how the pressure ramp-down in the last half of the shot pulled a far better result from single-origin coffees.

I had all but given up on, say, Coava's wonderful Centrals when pulled through the La Marzocco, getting superior (well, that is, drinkable) results only by carefully finessing the La Pavoni's lever. Most if the time I'd just punt and enjoy them through the Kone or French press instead.

I had no expectations that such a difference would be available in the espressos from these coffees simply by exercising my new-found ability to slow the second-half flow and thereby round off the aggressively sharp treble and bass corners.

I doubt I've got this consistently right yet. For example, if I just slam up to 9 bar and leave it there for the full shot, I'm not seeing a major acceleration in the flow toward the end - some perhaps, especially as the shot starts to lose its integrity, but nothing dramatic. So my "control" (un-profiled) shot may not be quite as different or rapid as the comparison that Jim describes. So I find myself ending the shot with a somewhat slower flow than the one with which I started (instead of keeping it the same), and judging more by the color progression than by time. FWIW, I'm also ignoring strict output weight parameters or expectations in favor of rolling with the shot's behavior on the fly.

These shot durations are consistently 10-15 seconds longer though, so Jim's comments give me hope I may at least have arrived at the ballpark. This logical explanation for what I've been tasting is also some relief - I may not be crazy after all, and with a sound model now in mind I'll be able to make some more progress.


#10: Post by erik82 »

What I noticed after experimenting with brew pressure in an E-61 HX machine was that taste got better (less harsh and more subtle) and flow improved a lot when setting the OPV to 8,5-9 bar from 11-12 bar. The difference between 9 bar and 10 bar gave less channeling and an overall better flow. Going up from 10 bar gave big differences in taste that I didn't like at all. These are my findings.

I currently have a Strega and agree with Jim on the differences a declining pressure profile gives. I'll never go back to a E-61 and am in love with the shots of the Strega.