Steam wand pressure shapes milk foam more than technique does

Put a thermometer in the jug the next time milk goes under the wand. Most machines at home get stopped well past the point where the foam is at its best, and the temperature is usually the last thing anyone checks. A steam wand heats and aerates in the same motion, which makes it the one part of the machine doing two jobs that pull against each other. Dairy scientists have measured both jobs, and the numbers have less to do with wrist technique than cafe folklore suggests.

The gear question underneath is a fair one. How much of the texture in the cup comes from the hardware, and how much from the hand holding the jug?

What the wand does to milk, and in what order

Steam pushed into cold milk does two things at once. It condenses, which dumps heat into the liquid fast, and it drags air down below the surface, which is where milk foam begins. Stop too early and there is air but no heat. Stop too late and the heat has arrived while the air has nowhere useful to go.

A 2015 study in the Journal of Food Engineering pulled those apart by building foams at steam pressures from 100 to 280 kPa gauge across three nozzle designs: an ejector, a plunging jet and a confined jet. Same milk every time. Only the hardware moved.

Pressure shifted the foam, which nobody would find surprising. The less obvious part is that the net effect depended on which nozzle was doing the injecting.

Where milk frothing actually peaks

A 2022 paper in the International Dairy Journal ran the steaming function of an espresso machine through final temperatures from 30 °C to 80 °C and measured stability, viscosity and bubble size distribution at each stop. The optimum for milk frothing came out between 50 °C and 60 °C. Past that the foam did not keep improving. It fell off.

An older result lines up with it. Work published in 2008 tracked foaming right across 5 °C to 85 °C and found that skim milk foams were most stable when they were formed at 45 °C, cooler than almost anyone serves. Neither figure is an industry standard, and neither was measured on plant milk. They are two sets of readings on dairy milk, and they point the same way: the hot end of the dial is not where the good foam lives.

Why the same steam pressure behaves differently on two machines

Nozzle geometry decides what a given pressure turns into. The number of holes, their diameter and the angle they sit at set how fast the steam leaves and how much air it carries down with it. Two wands fed from the same boiler at the same reading can make quite different foam, which is why a routine copied off someone else's machine tends to fall apart on your own.

The pump side of the machine teaches the same lesson. 9 bar became the espresso standard through habit and hardware rather than physics, and the reasoning behind that number reads well next to this. A figure on a spec sheet is a starting point, not a result.

Fat and protein pull in opposite directions

A 2015 image analysis study steam frothed milks at fat contents of 0.5%, 0.9%, 1.5%, 2.6% and 3.5%, and at protein contents of 3.0%, 3.4% and 3.8%. More protein delayed the moment a distinct foam layer appeared and slightly raised the volume. More fat cut foam stability, though it also reduced drainage, so what survived held its liquid better.

The 2008 work found the same tension from another angle. Milk fat hurt foam formation and stability in whole milk, worst of all across 15 °C to 45 °C, and bubbles in whole milk foams came out smaller while rupturing more through coalescence. Homogenisation and UHT treatment both softened the penalty. Temperature is doing as much work here as the carton is, which is why brewing temperature gets so much attention on the filter side too.

What to change on the machine, and what to leave alone

Three things are worth doing. Find out where the machine actually stops, with a thermometer rather than a guess. Aim for a frothing temperature between 50 °C and 60 °C instead of the point where the jug gets too hot to hold. And get the air in during the first seconds, because once the milk is hot the window for building microfoam has shut.

Milk choice is worth one experiment. In the 2022 work, foams from pasteurised milk tended to be more stable than the UHT equivalents, and the bubbles in the pasteurised foam were better distributed, especially at 60 °C. Tasters in the same study told the two apart by aroma and flavour, from moderate to strong.

What is not worth chasing is the swirl. The hardware sets the range and the thermometer sets the result. Most of what looks like skill is someone who has learned the timing of one particular machine.

Looking for a coffee that holds its shape under milk?

A milk drink flatters a thin coffee and exposes a dull one. Santo Café roasts a single origin Mexican arabica with enough body to read through the foam rather than disappear under it, which is the whole point of stopping the jug at 60 °C. Worth a try on the next flat white.