Where the 500 comes from
The formula every hydronic text repeats is BTU/h equals 500 times GPM times temperature drop, and it is worth knowing that 500 is three numbers multiplied together rather than a constant handed down. Sixty minutes in an hour, 8.33 pounds in a gallon of water, and one BTU per pound per degree Fahrenheit for water. Sixty times 8.33 times one is 499.8, which everyone rounds to 500. The rounding costs 0.04 percent and nobody minds.
What matters is that the second and third of those numbers are properties of the fluid. Put propylene glycol in the loop and the gallon weighs a little more while the specific heat drops well below one, and the product of the three falls. A lower constant means more gallons per minute to move the same heat at the same temperature drop, which means a larger circulator, more velocity in the same tube, and more head. This page takes the two fluid properties as inputs rather than assuming them, because the correct values are in the fluid manufacturer data at your concentration and your operating temperature, and they change with both.
The number that is not 500 and never was: 1.08 is the equivalent constant for air, built from the density and specific heat of air instead of water. Swapping the two is a factor of more than four hundred and it happens more often than it should, usually when someone reaches for a formula they half remember from the duct side of the job.
The trade between drop and flow
Load, flow and temperature drop are one equation with three terms, so fixing any two fixes the third. Design a circuit around a large temperature drop and it needs less flow, which means smaller pipe, a smaller circulator and less pumping energy. Design it around a small drop and the emitters see more uniform water along the circuit, which matters when several rooms are on one loop and the last one is getting whatever the first one returned.
| Temperature drop | Flow for 60,000 BTU/h on water | Velocity in a 0.811 in bore |
|---|---|---|
| 10°F | 12.00 GPM | 7.46 ft/s |
| 15°F | 8.00 GPM | 4.97 ft/s |
| 20°F | 6.00 GPM | 3.73 ft/s |
| 25°F | 4.80 GPM | 2.98 ft/s |
| 30°F | 4.00 GPM | 2.49 ft/s |
Doubling the design drop halves the flow and halves the velocity, and since friction loss climbs much faster than flow does, it cuts the head needed by far more than half. That is the argument for a large drop. The argument against it is that the emitters at the end of a long circuit are working with cooler water, and how much output they lose to that is exactly the calculation on the radiator output calculator and the baseboard length calculator. Neither side wins in general; it is a design choice with consequences on both sides of it.
Velocity, and why the page will not name a limit
The velocity arithmetic is simple: 0.4085 times GPM divided by the square of the bore in inches gives feet per second. The 0.4085 converts gallons per minute and square inches into feet per second, and the only trap in it is using the nominal size instead of the actual bore. Nominal three-quarter inch tube is not 0.75 inches inside in any material, and PEX in particular has a noticeably smaller bore than copper of the same nominal size, which changes velocity by a large factor because the diameter is squared.
What velocity is acceptable is a different question, and this page takes it as an input rather than supplying it. The constraints behind it are noise, erosion of the tube wall over years, air entrainment behaviour and, at the low end, whether the flow will carry air bubbles to the separator instead of parking them at a high point. The numbers that answer those come from the tubing manufacturer and from the design guidance you are working to, and they differ by material.
For the rest of the pipe problem — friction loss over the actual run, fittings, and the head the circulator has to make — use the water flow pressure drop calculator, which has the bore tables for copper, PEX, CPVC, PVC and steel. To see the same flow split among zones, the zone flow split calculator. For the volume of water the pipework itself holds, which matters for the buffer question, the pipe volume calculator.
Questions people ask
What is the formula for GPM in a hydronic system?
Flow in gallons per minute equals the load in BTU per hour divided by the product of the temperature drop and a fluid constant. For water that constant is 60 minutes times 8.33 pounds per gallon times a specific heat of 1.0, which is 499.8 and is universally rounded to 500. So 60,000 BTU/h at a 20 degree drop is 60,000 divided by 500 times 20, which is 6 GPM. If the system runs glycol, replace 8.33 and 1.0 with the values from the fluid data sheet at your concentration and temperature.
How much extra flow does glycol need?
It depends entirely on the concentration and the temperature, which is why this page asks for the properties instead of assuming a percentage. The mechanism is that the constant is proportional to weight per gallon times specific heat, and glycol mixes trade a slightly higher weight for a substantially lower specific heat, so the product falls. Enter the two values from the data sheet and the page reports the penalty directly as a percentage against water. The second effect, which this page does not compute, is viscosity: a glycol mix is thicker, so the same flow costs more head, and that correction comes from the fluid tables too.
Is 500 or 1.08 the right constant?
Five hundred is water and 1.08 is air. They come from the same derivation — sixty minutes times a density times a specific heat — applied to two very different fluids. Air at standard conditions weighs about 0.075 pounds per cubic foot with a specific heat near 0.24, and sixty times those gives 1.08, used with CFM instead of GPM. Using the air constant on a water circuit understates the heat a given flow carries by more than four hundred times, and the mistake usually shows up when someone is working on both sides of the same job in one afternoon.
What temperature drop should I design for?
It is a choice rather than a rule, and the trade is real in both directions. A larger drop means less flow, smaller pipe and much less pumping head, since friction rises roughly as flow to the power 1.85. A smaller drop means the water reaching the far end of a circuit is closer to supply temperature, so emitters late in the loop perform closer to those early in it. Systems built around low-temperature emitters often use a smaller drop for exactly that reason, because at low supply temperatures there is less room to give any away. Whatever you choose, keep it consistent with the temperature the emitters were sized at, since the emitter rating uses the average of supply and return.
Why is my measured temperature drop different from the design figure?
Because the drop is an output, not a setting. It is whatever the load and the actual flow produce. At part load, with the same flow, the drop is smaller, in proportion to the load. If the drop is much larger than designed at full load, the circuit is moving less water than intended, and the usual reasons are a partly closed balancing valve, air in a high point, a dirty strainer, or a circulator working against more head than anyone calculated. If the drop is much smaller, the circuit is moving more water than intended, which is common when a fixed-speed circulator was chosen with margin and the balancing was never done.