There are two gallons
The US gallon is 3.785411784 litres. The Imperial gallon is 4.54609 litres. Both are exact by law, both are called “a gallon”, and the Imperial one is 20.095% larger.
They were never the same. The US kept the old English wine gallon of 231 cubic inches after independence; Britain replaced its own tangle of gallons in 1824 with a new one defined by the volume of ten pounds of water. So the divergence is two centuries old and neither side is going to move.
The same number, read under each definition| Written | If US | If Imperial | Difference |
|---|
| 1 GPM | 3.79 L/min | 4.55 L/min | 0.76 L/min |
|---|
| 2.5 GPM | 9.46 L/min | 11.37 L/min | 1.90 L/min |
|---|
| 10 GPM | 37.85 L/min | 45.46 L/min | 7.61 L/min |
|---|
| 50 GPM | 189.27 L/min | 227.30 L/min | 38.03 L/min |
|---|
The 2.5 GPM row is the one that matters most often. That is the US federal maximum for a shower head, and it is 9.46 litres a minute. Read as Imperial it becomes 11.37 — a fifth more water, which is the difference between meeting a water-efficiency rule and missing it.
Where you meet each: American pump curves, irrigation and plumbing are in US gallons. Older British and Commonwealth equipment — boilers, burners, legacy pump data — is in Imperial. Modern British practice has largely moved to litres, which is the safest thing to convert to, because “litres” is not ambiguous. If a figure in gallons does not say which kind, it is not yet a specification.
A cubic foot of air is not a quantity of air
For water this whole section is unnecessary: a litre of water is a litre of water at any sane pressure. Gas is different, and the difference is large enough to buy the wrong compressor with.
Compress air and you fit more of it into the same space. So a cubic foot at 100 psi contains far more air than a cubic foot at atmospheric pressure — the same volume, a very different quantity.
What 100 standard cubic feet per minute occupies at working pressure| Gauge pressure | Actual flow | Compression ratio |
|---|
| 0 psig | 100 ACFM | 1.0× |
|---|
| 50 psig | 22.7 ACFM | 4.4× |
|---|
| 100 psig | 12.8 ACFM | 7.8× |
|---|
| 150 psig | 8.9 ACFM | 11.2× |
|---|
So the two figures have names. SCFM is the flow referred back to a standard condition — a real quantity of air, and what you compare compressors on. ACFM is the volume actually passing a point at the working pressure and temperature — what a duct has to carry and what an anemometer measures.
A compressor advertised at “100 CFM” with no qualifier is unspecified by a factor approaching eight. Reputable manufacturers state SCFM at a named discharge pressure; the figure to compare is always at the pressure you will actually run at, because output falls as pressure rises.
Temperature does the same thing more gently. At constant pressure, air at 200 °C occupies 61% more volume than at 20 °C, which is why hot-side duct sizing and cold-side duct sizing are different calculations for the same mass of air.
“Standard” is four different things
Having established that a gas volume needs a reference condition, the obvious question is which one. There is no single answer, and the disagreements are large enough to matter commercially.
Reference conditions in common use, all at 101.325 kPa| Condition | Temperature | Where |
|---|
| Normal (Nm³) | 0 °C | Europe and India, and the chemistry convention |
|---|
| Standard, 15 °C | 15 °C | ISO 13443, natural gas metering |
|---|
| Standard, 60 °F | 15.56 °C | US oil and gas |
|---|
| Standard, 20 °C | 20 °C | US compressed air (CAGI, ASME) |
|---|
A normal cubic metre and a standard cubic metre at 20 °C differ by 7.32%, because gas volume scales with absolute temperature and 293.15 K over 273.15 K is 1.0732. On a gas contract that is a seven percent billing error, in whichever direction favours whoever wrote the specification.
And 60 °F is not 15 °C — it is 15.56 °C. The two look interchangeable and are 0.19% apart, which is small until it is applied to a pipeline’s annual throughput.
The practical rule is the same as for gallons: a gas volume without its reference condition is not a measurement. Nm³/h, Sm³/h and SCFM all mean “a quantity of gas”, but only once the subscript is honoured.
Volume is not mass
Kilograms per hour is a different quantity from litres per minute, and no factor connects them without knowing what is flowing.
Ten cubic metres an hour, as mass| Fluid | Density | Mass flow |
|---|
| Seawater | 1025 kg/m³ | 10,250 kg/h |
|---|
| Water at 20 °C | 998.2 kg/m³ | 9,982 kg/h |
|---|
| Diesel | 832 kg/m³ | 8,320 kg/h |
|---|
| Petrol | 745 kg/m³ | 7,450 kg/h |
|---|
| Air at 20 °C | 1.204 kg/m³ | 12 kg/h |
|---|
Water against air is a factor of 829. Even among liquids the spread is a third, which is why fuel is bought by volume and burned by mass, and why a fuel volume figure without a temperature is worth arguing about — diesel expands roughly 0.08% per degree, so a tanker loaded warm and delivered cold is short by a measurable amount of mass while the volume reads correctly.
The tool asks which fluid rather than assuming water. That is a deliberate extra click: quietly defaulting to water gives an answer that is wrong by up to three orders of magnitude and looks exactly like a right one.
What flows where, in familiar numbers
Flow rates span an enormous range and the units obscure it. A few anchors make a converted figure easy to sanity-check.
Everyday and industrial flows for comparison| Something | L/min | Other units |
|---|
| A dripping tap | 0.02 | about 30 L a day |
|---|
| A bathroom tap | 6 | 1.6 US GPM |
|---|
| An efficient shower | 9.5 | 2.5 US GPM |
|---|
| A garden hose | 20 | 5.3 US GPM |
|---|
| A bathroom extractor fan | 1,400 | 85 m³/h, 50 CFM |
|---|
| A fire hydrant | 3,800 | 1,000 US GPM |
|---|
| A small river | 600,000 | 10 m³/s |
|---|
Two things fall out of that table. Air moves in numbers ten to a hundred times larger than water for the same duty, which is why ventilation is quoted in m³/h or CFM while plumbing is in L/min — and why mixing the two habits produces answers that are obviously wrong once you notice the scale.
And a dripping tap at 0.02 L/min sounds negligible until it is a rate: thirty litres a day, eleven cubic metres a year. Flow rates multiply by time, which is the whole reason they are worth converting carefully.
Sources and methodology
Both gallons are exact by law rather than by measurement, so the 20.095% gap is a fixed figure and the references establish it. The gas behaviour is the ideal gas law, which is why the reference conditions matter more than the arithmetic — and those are conventions set by standards bodies that did not agree with each other.