Pressure and flow in espresso

One sets the other

Sourced

A bed of ground coffee resists the water pushed through it. For a given bed, more pressure produces more flow, up to a point; equally, demanding more flow requires more pressure. The authors of one study put it in terms of the standard law for flow through porous material: at otherwise constant conditions, an increasing flow rate corresponded directly to an increasing pressure. A later study that varied pressure from 1 to 12 bar found that this proportion held only below about 5 bar; above that the bed was squeezed tighter by the water and the flow levelled off. That finding is for the steady flow about 100 seconds into a brew, after a normal shot would have ended; during the first half minute the flow was still rising at every pressure but the lowest, and the cup filled fastest at about 5 bar.

A typical pump machine applies a pressure and lets the flow be whatever the bed allows. That is why grind dominates espresso: a finer bed resists more, so the same pressure yields a slower shot. The machine in that study worked the other way round, holding a chosen flow rate and letting the pressure follow, which is unusual and is the reason it could vary flow cleanly.

The resistance is the barista's doing. Dose, grind and how evenly the bed is prepared decide it, and it changes during the shot as the coffee swells and gives up material.

The pressure a machine reports is also not the pressure on the coffee. The same later study put a second sensor beneath the basket and found that the pipework, valves and group between the pump and the bed take a share of the pressure that grows with the flow rate. Two machines showing the same figure need not be pressing on their coffee equally hard, and a lever machine, whose spring or the operator's arm supplies a pressure that falls away through the shot, is different again. How a bed behaves as a porous material is the subject of the page on espresso puck physics.

What changing the flow did

Sourced

The study collected each shot in ten consecutive fractions and measured caffeine, two other compounds and total dissolved solids in each, at flow rates from 1 to 3 millilitres per second, three grind settings and temperatures from 80 to 98 degrees Celsius. Of the three variables, flow rate had the strongest effect on what ended up in the cup. Raising it reduced the amount of each component extracted, with caffeine the least affected.

The effect of flow was more pronounced at finer grinds and higher temperatures. And all of it, the authors say, was minor compared with the effect of how much liquid was drawn in total: the difference between a short shot and a long one outweighed everything else they varied.

They are careful about the limits. Their data, they write, allow qualitative statements and not quantitative ones. It was one coffee, one grinder and one machine, and nobody tasted the shots.

Pressure as a sign of what the bed is doing

Sourced

The study reports the pressure measured at its three grind settings as 3.8, 7.4 and 9.3 bar, from coarsest to finest. By the authors' account the measured particle size distributions at those settings were nearly identical, and one summary measure of them did not differ significantly. A difference too small to show clearly in the measurement of the grounds went with a pressure more than twice as high. Pressure also rose with flow, far more steeply at the finest setting than at the coarsest.

The authors connect this to an earlier study in which extraction fell at the finest grinds, which that paper's authors attributed to parts of the bed clogging; they say their pressures agree with that hypothesis. A bed that needs more pressure to keep a set flow is what clogging would look like from outside. It is an interpretation that two groups have reached, and neither observed the inside of the bed directly.

Familiar rules, and what stands behind them

Accounts differ

Nine bar

The study quoted here describes a commonly recommended method as using a pump at nine bar, give or take two, and says such ranges leave much to the barista because the effects are not fully understood. It is a convention of long standing; a physics paper on espresso flow records that it is often traced to the force the first lever machines of the late 1940s could deliver. No study read in full shows nine bar to be better than another pressure. A review of espresso aroma reports, from studies CoffeeHQ has not opened, that shots at nine bar had a firm foam and intense aroma and that raising the pressure to eleven increased bitterness and astringency; that is one line of evidence at second hand. The other espresso study read here ran its shots at six, because nine clogged the bed at fine settings, and its model predicted more extraction at lower pressure.

Pre-infusion

Wetting the bed at low pressure before the full pressure arrives is standard on many machines; the study's machine filled the basket first and switched over once pressure passed 2.5 bar. Brewers credit it with a more even shot, and filter brewers call the equivalent step a bloom. A mathematical model of espresso concludes that a significant part of what ends up in the cup is extracted while the bed is first being wetted, which is a reason to think the stage matters. One group has watched that stage by X-ray on a home machine: water took about seven seconds to reach the bottom of a finely ground bed, and by the group's model the pressure on the bed built up gradually as the empty space above the coffee filled. On that machine a gentle start is what the pump and the space above the bed produce with no pre-infusion setting at all. No test of what pre-infusion does to evenness or to taste was read.

Pressure and flow profiling

Machines that vary pressure or flow through a shot are sold on the control they offer. The practice is real and widely used; what a given profile does to taste has not been established in anything read here. What has been measured is that neither quantity stays put when the other is held. In the study that set a constant pressure and watched the flow, flow went on rising through the shot as the bed gave up material; in the study that set a constant flow, pressure was what moved. A lever machine's falling pressure is a third case, and no study read compares it with either.

Faster, coarser shots

Sometimes called turbo shots. They follow a study that found extraction peaking at an intermediate grind and proposed coarser grinding with a smaller dose, fifteen grams in place of twenty, to make shots more reproducible and use less coffee. Its authors say the result may be a very fast shot, weaker, with a different flavour, and that the slower, partly clogged shots tasted more complex. The proposal is about consistency and waste, not about taste.

Reading a shot by its flow

CoffeeHQ explanation

Runs fast and tastes thin or sharp

Usually because: The bed offers little resistance, commonly because the grind is too coarse or the dose too small for the basket.

Change first: Grind finer, keeping dose and yield the same.

Shot runs too fast

Barely drips, or stops

Usually because: The bed resists too much, commonly because the grind is too fine.

Change first: Grind coarser, keeping dose and yield the same.

Shot runs too slow

Starts slowly, then gushes

Usually because: Possibly water breaking through part of the bed and bypassing the rest.

Change first: Look at how the bed was prepared before changing the grind.

Channelling

What lies beyond this page

CoffeeHQ explanation

It gives no pressure, flow rate or shot time as correct. It does not describe pump types, lever machines or how a particular machine regulates pressure; a machine's own manual covers that. It makes no claim about what pre-infusion or any profile does to flavour. And it does not carry the one study's figures over to other grinders and machines, where the same settings would give different pressures.

What to do next

• Espresso dial-in assistant — Takes dose, yield, time and taste and suggests one change, keeping flow and taste apart.
• What your espresso machine can do — Which machines offer pre-infusion or pressure control, and how to check your own.
• Advanced espresso lesson — Pre-infusion, pressure, flow and measurement, with an exercise.
• Grinding and particle size — The bed's resistance starts with the grind.
• Espresso puck physics — The bed as a porous medium: compaction, fines, uneven flow, tamping and what channelling means.