Why a Pressure Cooker's Minimum Liquid Is Not Steam

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

Because the liquid at the bottom of the pot is not there to make steam, and the alert is telling you about the only job it does have. Work out how much steam a six quart pressure cooker actually holds at full pressure and the number is 4.82 grams. Less than a teaspoon. Everything else in that cup of water is doing something else entirely, and when it stops doing it, the machine complains.

This also answers the quieter question that sits behind every pressure cooker recipe, which is how much water to put in for pasta or rice. The honest answer has two terms and neither of them is steam.

The entire steam charge, weighed

Six US quarts is 5.6781 liters. Put a liter of water in the bottom and the headspace above it is 4.678 liters.

Our page on why steam cannot get hotter than 100 degrees establishes the operating band and this page takes its figure rather than re-deriving it: 11.6 pounds per square inch of gauge pressure, which is 79.98 kilopascals, is the top of the usual electric range. Add atmospheric and the absolute pressure is 181.31 kilopascals. Water boils there at 117.1 degrees Celsius, and saturated steam at that pressure has a density of 1.030 kilograms per cubic meter.

So the mass of steam needed to fill the entire headspace is 4.678 liters times 1.030 kilograms per cubic meter, which is 4.82 grams. A US teaspoon is 4.93 milliliters. The whole steam inventory of the appliance is 98 percent of one teaspoon of water, and its latent heat is 4.82 grams times 2,215 kilojoules per kilogram, which is 10.67 kilojoules.

That is the number that makes the rest of the page obvious. Whatever the minimum liquid is for, it is not for that.

What the cup of water is actually buying

An Instant Pot Pro 6 Qt 10-in-1 electric pressure cooker publishes 1200 watts, 28 one-touch programs, a 6 quart stainless inner pot and a 120 volt 60 hertz supply, and that wattage is what turns this from a thermodynamics exercise into a clock. Lay out the energy needed to get the thing to pressure with one liter of water and no food in it.

Heating the water from 20 to 117.1 degrees is one kilogram times 4.186 kilojoules per kilogram per kelvin times 97.1 kelvin, so 406.5 kilojoules. Heating the inner pot, taking a mass of 1.1 kilograms and a specific heat of 0.50 for stainless, is 53.4 kilojoules. Making the steam is 10.67 kilojoules. The total is 470.6 kilojoules, and the split is 86.4 percent water, 11.3 percent pot, and 2.3 percent steam.

At 1200 watts with 85 percent of it reaching the contents rather than the housing, that is 461 seconds, or 7.69 minutes.

Now price a cup. Two hundred and thirty-six point six milliliters of water is 96.2 kilojoules of sensible heat, which at the same power is 94 seconds. One cup of water costs you a minute and a half of preheating and supplies nine times more energy than the entire steam charge requires. Per 100 milliliters it is 40 seconds.

And here is the part that is genuinely backwards from the way this gets explained. Go from one liter to two. The water now takes 813.0 kilojoules instead of 406.5 and the total climbs to 874.8 kilojoules, which is 14.29 minutes. The time to pressure has risen 86 percent. Meanwhile the headspace has shrunk from 4.678 liters to 3.678, so the steam the cooker has to make has fallen from 4.82 grams to 3.79, a drop of 21.4 percent.

Adding water reduces the steam requirement and lengthens the time to meet it. The second liter removes work from the one stage people believe it is there to supply.

So what is the liquid for

Two things, and they are the two things that actually set how much to use.

It gets absorbed. This is the whole of the pasta answer. Dry pasta takes up roughly 1.5 grams of water per gram of itself on the way to being cooked, so 500 grams of dry pasta needs about 750 grams of water taken up into the food and gone from the pot. That is 3.2 times the usual one cup minimum and 156 times the steam charge. Rice is the same argument with a different ratio. If you put in one cup for a pound of pasta, the pasta wins the argument with the base of the pot, and the alert is the pot telling you so.

It keeps the base wet. Our page on why a pressure cooker is tall and its pot is not works out the geometry this needs and this page takes those figures: the pot's base is 0.0356 square meters and the flux through it at 1200 watts is 33.7 kilowatts per square meter. That element does not modulate the way a stovetop flame does. The only thing holding the base at 117 degrees instead of running away is liquid water touching it, because boiling water pins a surface to its saturation temperature and nothing else in the pot can.

A two millimeter film of water over that base is 71.2 milliliters. At 1200 watts of pure evaporation it would be gone in 131 seconds, which sounds alarming until you remember the lid is shut. In a sealed vessel at steady pressure, net evaporation is zero: every gram that leaves the surface condenses on the lid and runs back. So the film does not go anywhere, and that is why a pressure cooker never needs topping up and an open pot does.

Put a number on the contrast. An open pot boiling at 2000 watts sheds 2000 times 600 seconds divided by 2,257 kilojoules per kilogram, which is 532 grams of water in ten minutes. Half a liter, straight up the extractor. The sealed cooker sheds none, which means every gram of water you add to a pressure cooker ends up in the dish. That, and not the steam, is the reason the advice for pressure-cooked pasta is a fraction of the water you would boil it in.

The published cook time that leaves out the longer half

A Instant Pot Duo 6 Qt 7-in-1 electric pressure cooker publishes 13 one-touch smart programs, an 18/8 stainless pot with a tri-ply bottom and six servings of capacity, and its record is unusually frank about what the programs are: they "act mainly as preset timers for specific foods" and are "not fully automatic, so adjust time by quantity and food type". That is exactly right, and the arithmetic above says why the quantity term is so strong. A program is a timer that starts after pressure is reached, and the thing it does not time is the 7.69 minutes.

A Ninja HyperHeat 9-in-1 pressure cooker is the listing that does the honest thing here, and it deserves saying. It publishes 6.5 quarts, white rice "in as little as 8 minutes" and quinoa in as little as 5, and then states plainly in its own drawbacks that "Rice and grain times exclude pressure build/release". Most of this category leaves that unsaid.

What it cannot do is tell you how big the excluded part is, because it publishes no wattage. Borrow the 1200 watts from the machine above and run a rice-sized charge in a 6.5 quart pot, two cups of rice and two cups of water, and the preheat comes to about 5.2 minutes. So an 8 minute published figure is roughly 61 percent of a 13.2 minute reality, and the missing 39 percent is the part that scales with how much water you poured in. The listing's honesty about the exclusion is worth more than the 8, and the number that would make the 8 usable is the wattage it does not give.

A record that admits the gap, and one that cannot

A Cosori 6QT pressure cooker is the clearest case of a listing knowing what it is missing. It publishes 13 modes against a 9-in-1 badge, which is a pleasant direction for a discrepancy to run, a ceramic-coated inner pot, 12 safety features, a steam release angled 30 degrees backward and a pot gap widened by half an inch. Then its own drawbacks say: "No explicit mention of pressure levels or wattage; check listing for details."

That single sentence is the most useful thing in the category, because it is true of almost all of it. Not one of the seven pressure cookers we carry publishes an operating pressure in any unit. Only two publish a wattage, and the two stovetop models cannot. None publishes an inner pot mass, which is 11.3 percent of the preheat above and had to be assumed. And not one of the seven publishes a minimum liquid volume, which is the single number every recipe written for these machines depends on and the subject of the question this page answers.

An Instant Pot Duo 7-in-1 Mini publishes 3 quarts, 700 watts, 11 programs and a stainless tri-ply pot, and it is worth running the inventory on because the result is so small. Three quarts is 2.839 liters, and with half a liter of water the headspace is 2.339 liters, which is 2.41 grams of steam. Under half a teaspoon, for an entire appliance at full pressure. Its preheat on the same assumptions is 242.6 kilojoules, or 6.8 minutes at 700 watts, and the steam is 2.2 percent of it, the same fraction as the larger machine. The share does not change with size, because both terms scale with volume.

On the 700 watt figure itself, that record lists as a drawback that "700-watt power may heat slower than higher-wattage models". Our geometry page has already shown why that is inverted at a proportionate fill, and the finding stands rather than being repeated here. What the inventory adds is the smaller point that the Mini's steam charge is half the larger machine's while its element is 58 percent of the size, so on the steam stage alone the small machine is ahead.

Three speed claims that cannot be ranked

Four of the seven machines make a speed claim and they are not in the same units, which is a problem that gets worse the closer you look.

Three of them, including both Instant Pots above, publish "up to 70% faster than slow cooking". The Ninja publishes "up to 2 times faster than traditional slow cooking". Those sound like the same kind of statement. They are not, because "70 percent faster" has two ordinary readings and nothing on any listing says which is meant.

Read it as a 70 percent reduction in time and the speed ratio is one divided by 0.30, which is 3.333. On that reading the Instant Pots claim 3.333 and the Ninja claims 2, so the Instant Pots win by 67 percent.

Read it as a 70 percent increase in speed and the ratio is 1.70, with time falling 41.2 percent. On that reading the Ninja's 2 wins by 18 percent.

The ranking reverses. One phrasing, two readings, opposite answers, on a claim that three of our seven machines carry in identical words. And then a Presto 01365 6-quart stainless steel pressure cooker publishes "3 to 10 times faster than traditional methods", which is a third baseline: traditional methods means boiling and braising, not slow cooking, so its range cannot be compared with either of the others even after you pick a reading.

What the physics allows is more than any of them claim, which is worth saying because it means the claims are vague rather than inflated. Our page on why a slow cooker low setting is a wattage puts slow cooking at 85 to 95 degrees, and the steam page gives the conventional Q10 band of 2 to 4 for food reactions. From 90 degrees to 117.1 is 27.1 kelvin, so the rate ratio is somewhere between 2 to the power 2.71 and 4 to the power 2.71, which is 6.5 to 42.8. Against boiling at 100 degrees the gap is 17.1 kelvin and the ratio is 3.3 to 10.7, which is almost exactly the Presto's published 3 to 10. The stovetop model with no electronics has the most defensible number on the shelf.

The stovetop case, and one claim nobody can check

The Presto is useful here for a second reason. It has no element, no board and no minimum liquid problem of the electric kind, because the heat arrives through the base from a burner you control, and its record notes it is "fully immersible and dishwasher safe" precisely because there is nothing inside to protect. A burner you can turn down is a different machine from a 33.7 kilowatt per square meter element you cannot, and the burn alert exists only on the appliance that cannot modulate.

A T-Fal Secure stainless steel stovetop pressure cooker is the same idea at 6.3 quarts with a 5-point security system, a rotating lock with an audible click, induction compatibility and two programs. It publishes "up to 2x faster than conventional methods", which is a fourth phrasing of the speed claim and the lowest number on the shelf for a machine that reaches the same temperature as every other one here.

It also publishes the one figure across all seven listings that cannot be checked in either direction: its vegetable program "preserves up to 80% of vitamin C in veggies". There is no baseline named, no cooking time, no vegetable and no comparison method, so there is no arrangement of numbers that would make it true or false. Every other claim on these seven pages is at least the right shape to be tested. This one is not, and it is the only percentage in the category that is about the food rather than the clock.

The honest limit, and the two numbers to take away

What is exact on this page is the inventory, and it is exact because it is a volume times a density. At 181.31 kilopascals absolute, saturation is 117.1 degrees, saturated steam weighs 1.030 kilograms per cubic meter and carries 2,215 kilojoules per kilogram. Multiply by a headspace and you have grams. Nothing in that depends on the appliance, the food or the recipe, which is why the 4.82 grams and the fact that it falls as you add liquid are the two results worth keeping.

What is representative rather than published is most of the clock. The inner pot mass of 1.1 kilograms is not stated by any of the seven, and it is 11.3 percent of the preheat. The 85 percent of element power reaching the contents is a reasonable figure for an appliance with a sleeved element and a plastic shell, and it scales every time on this page directly: at 95 percent the 7.69 minutes becomes 6.88, and at 70 percent it becomes 9.34. The specific heat of stainless at 0.50 and the 1.5 grams of water taken up per gram of dry pasta are both representative. So read the minutes as the right size and the ratios as firm, which is the same division as every other page here.

What this page deliberately does not compute is a cooking time for anything. That needs a rate constant for the specific food and the specific texture you want, and not one listing in this category publishes one, nor does any of them state a pressure that a rate could be attached to. The Q10 band of 2 to 4 is wide enough that a ten minute boil maps to anywhere between 56 seconds and three minutes under pressure, which is a factor of three, and quoting a single figure out of that range would be inventing precision. Nor has this page said anything about what is safe to cook or for how long, because that is not a question arithmetic on a headspace can answer.

And the practical residue. For pasta or rice in one of these, the water you need is the water the food will absorb, plus a shallow film to keep the base wet, and nothing at all for steam. Everything above that is 40 seconds of preheating per 100 milliliters and it stays in the pot, because a sealed vessel does not evaporate. If you are getting the burn alert, the fix is almost never more total water: it is water that is free and in contact with the base rather than locked up in something starchy sitting on it, which is why the standard advice to layer liquid first and not stir the sauce in works, and why it works on a 1200 watt machine more than on a burner. Our roundup of electric pressure cookers worth the counter space is the place to compare the machines themselves. Read the wattage first, because on two of the seven it is the only number that connects a published cook time to a real one.