How Many Bars of Pressure for Espresso? What 9, 15 and 20 Really Mean

Ask how many bars of pressure for espresso and you get two different numbers from the same manufacturer. Breville’s product page for the Barista Express says the machine brews at 9 bar. Three lines later it says the water is “delivered via our 15 bar Italian pump”. Both are true, and the gap between them is the whole story. One number is what the coffee sees. The other is what the pump could push against a blocked pipe. This guide sorts out which is which, how a machine holds the number, and what a 2026 fluid-dynamics paper says about whether 9 was ever the right target.

Reasoned. Pressure in espresso is a result, not an input you dial in from the box. The pump pushes, the coffee puck resists, and the number on the gauge is what those two agree on. That is why a machine rated at 15 bar can read 6 bar on a coarse grind and pin the needle on a fine one. Keep that in mind and most of the marketing around bar counts stops mattering. For where the pump sits in the full water path, see how an espresso machine works, stage by stage.

What a bar is, and why 9 keeps coming up

Documented. A bar is a metric pressure unit. NIST’s Guide for the Use of the International System of Units (Special Publication 811, Appendix B) defines it as exactly 100 kilopascals, and gives one pound-force per square inch as 6.894757 kilopascals. Run the arithmetic and one bar is about 14.5 psi, so 9 bar is roughly 130 psi. For scale, that is nine times the air pressure at sea level, which is close to 1 bar by definition. A car tire runs at a bit over 2 bar.

Documented. Nine is the number manufacturers put on paper when they describe the shot rather than the pump. Breville’s US product page for the Barista Express describes the brew cycle as “low pressure pre-infusion followed by 9 bar extraction”. La Marzocco’s Linea Mini page does not quote a bar figure at all; it lists “easily adjustable pump pressure” as a feature and leaves the target to the owner. Neither company claims that 9 is a law of nature. It is the setting their machines ship with or are tuned to, and it is what the pressure gauge zones on a Breville are drawn around.

Reasoned. Where 9 came from is a history question this page does not try to settle. What matters for a buyer is narrower: 9 bar is the figure the industry builds, tunes and labels around, so a machine that holds it steadily is doing the job the coffee was roasted and ground for.

15 bar vs 20 bar: what the pump rating describes

Documented. Breville’s own manual for the Barista Express is the clearest illustration that the pump number and the brew number are different things. The troubleshooting table explains a pulsing sound during extraction or steaming as “the normal operation of the 15 bar Italian pump”. The same manual’s gauge section says the needle should sit in the “ideal espresso zone” during extraction, and the product page puts that zone at 9 bar. So one document, one machine, two numbers, and the manufacturer never suggests the coffee sees 15.

Reasoned. The pump rating is a ceiling. A small vibratory pump, the type in nearly every entry-level machine, is specified by the maximum pressure it can develop when the outlet is nearly blocked. That is the 15 or 20 on the box. In use the water runs through a puck of ground coffee that offers resistance, and if the puck resists more than the brew target, the machine needs something to dump the surplus. That something is an over-pressure valve, usually a spring-loaded brass fitting that opens above a set point and routes water back to the tank. The spring, not the pump, decides the number at the group. A 20 bar pump behind a 9 bar valve brews at 9. So does a 15 bar pump behind the same valve.

Reasoned. This is also why “15 bar vs 20 bar” is not a useful comparison between two machines. Both pumps out-push a 9 bar valve with room to spare. What separates machines is whether the valve is set sensibly from the factory, whether it can be adjusted, and whether the machine holds temperature while it holds pressure. Those are the criteria behind what makes an espresso machine prosumer, and pump rating is not on the list.

How a machine holds the number: valve and gauge

Documented. Breville’s manual describes four gauge positions for the Barista Express. At the start of a shot the needle sits in a pre-infusion zone while the machine ramps pressure up gently. A needle anywhere in the solid grey zone during the shot means the espresso “has been extracted at the ideal pressure”. A needle in the lower zone means insufficient pressure, which the manual ties to water running through the coffee too fast, with the grind too coarse, too little coffee in the basket, or under-tamping as causes. A needle in the upper zone means too much pressure, tied to flow that is too slow: grind too fine, too much coffee, or over-tamping.

Reasoned. Read that list again and notice that none of the causes is the pump. On a machine with a fixed valve, the pressure reading is a symptom of the puck: the pump supplies a constant push, and the grind, dose and tamp set how much of it the coffee absorbs before the valve opens. That is why “my machine only reaches 6 bar” is usually a grind, dose or basket complaint, and occasionally a mechanical one, which we walk through in why your espresso machine is losing pressure. A needle that swings during the shot points the same way: water finding an easy path through the puck, which is the mechanism behind espresso channeling.

Is 9 bar better than 6, 12 or 15? What the fluid-dynamics study found

Documented. In June 2026, Physics of Fluids published “Under pressure: poroelastic regulation of flow in espresso brewing” by Radost Waszkiewicz, Maciej Lisicki and five co-authors (also posted to arXiv in December 2025). Working on a café-grade espresso machine, the team found that the interplay between the coffee bed’s elasticity and its porosity governs the long-time flow rate, and they built a minimal model of the resulting non-linear relationship between pressure and flow. In plain terms: past a certain point, pushing harder does not move water through the puck faster, because the puck compresses and closes its own pores. Gear Patrol’s July 2026 report on the paper put the onset of that behavior at roughly 5 bar, with flow flattening out well below the 9 bar standard. The abstract itself does not name a replacement pressure, and neither does this page.

Reasoned. That result matches what the gauge zones in the Breville manual already imply. Above the ideal zone the manual predicts a shot that is “very dark and bitter” with uneven crema, which is what you expect when a fine puck compacts under pressure and water crawls through it. Below the zone it predicts watery, thin coffee, which is what you expect when the puck never built resistance in the first place. The study puts physics behind the picture: pressure helps up to the point where the bed starts behaving like a sponge being squeezed, and after that extra bar is mostly extra compaction.

Reasoned. So is 9 bar “better” than 6? For most owners the honest answer is that 9 is the setting your baskets and recipes were developed around, and changing it means re-learning them. Owners of machines with adjustable pressure do brew lower and get good results, which is consistent with the paper. What nobody serious argues for is 15 or 20 at the puck. That number only ever described the pump.

Why machines start below 9: pre-infusion

Documented. Breville’s manual describes the Barista Express’s low-pressure pre-infusion as a phase that “gradually increases water pressure to gently expand grinds for an even extraction before increasing to full pressure”, and lets the user hold the shot button to extend it manually. La Marzocco’s Linea Mini page describes a two-valve pre-infusion system “for soft saturation” with a timeout that can be set up to 99 seconds. Two manufacturers at opposite ends of the home market, same idea: wet the puck at low pressure before asking it to resist 9.

Reasoned. Pre-infusion exists because of the compaction effect described above. A dry puck hit with full pressure compacts unevenly and channels; a puck that has been saturated first expands and settles before the load comes on. On an E61 group the same job is done mechanically by a spring-loaded chamber, which is covered in how the E61 group head works. The bar count during pre-infusion is low on purpose, and a gauge dipping into its pre-infusion zone at the start of a shot is the machine working as designed, not a fault.

Two questions the search results keep asking

Is 20 bars of pressure good for espresso?

Reasoned. As a pump rating it is fine and changes nothing; as a brew pressure it would be well past the point where the puck stops cooperating. No manufacturer documentation we opened describes brewing at 20 bar. A machine sold as 20 bar is a machine with a pump that can reach 20 bar if nothing is in the way. What matters is where its valve is set, and the box rarely says.

Is 10 or 12 bar too much?

Reasoned. Slightly above 9 is where many factory valves land, and the shot will not fall apart. Well above it, the Breville gauge logic applies: slow flow and a darker, more bitter cup unless you coarsen the grind. If the gauge reads high on every shot regardless of grind, the valve set point is the thing to have checked, which is a technician’s job on a sealed machine. Tamping force, by the way, is a different quantity: it shifts the puck’s resistance a little but adds no bars, as our tamping guide explains.

The short version

Reasoned. Espresso is brewed at about 9 bar, roughly 130 psi, and that is the figure manufacturers put on the shot. The 15 or 20 bar on the box is the pump’s ceiling, and a valve inside throws the extra away. On a fixed-valve machine, the gauge reading is set by your grind, dose and tamp, not by the pump. The 2026 study says flow through the puck stops responding to extra pressure well before 9 bar, which argues for consistency at whatever your machine holds rather than for chasing higher numbers. Compare two machines by bar count and you are comparing pumps, not shots. This page is built on manufacturer documentation and a published study, not a bench test we ran ourselves; see how we evaluate gear we haven’t bench-tested.