SizingKitPipe & pumps
NPSH calculator
NPSH available is what the system delivers to the impeller eye above the pressure at which the liquid boils, and this page returns it as the four terms it is built from — surface pressure, vapor pressure, static column and suction friction — in feet and meters at once. Enter the NPSH required off the pump's own curve and it reports the margin, the ratio and how far above the liquid the pump may sit while keeping it. It costs nothing and asks for no account, and it sweeps the same system from 50 °F to 212 °F so the temperature at which the margin disappears is a row in a table rather than a surprise.
- 100% free
- No signup
- NPSHa in four terms
- Vapor pressure 33–212 °F
- Barometric to 20,000 ft
- Printable temperature sweep
The suction side
Everything between the liquid surface and the impeller eye. What the pump has to make once the liquid is inside it belongs to the pump head calculator.
Outside 0.01 to 1,000,000 gpm. NPSH required is read off the curve at a flow, so a flow of zero has nothing to check.
68 °F is 20 °C. Prefilled at 68 °F, the 20 °C anchor row of the water table — ρ 62.316 lb/ft³, vapor pressure 0.3392 psia.
Barometric pressure 14.7 psia there, against 14.7 psia at sea level.
A deaerator or a condenser hotwell is the hard case: its liquid is already at its own boiling point, so the pressure term and the vapor term cancel exactly and every foot of NPSH has to come from the static column above the pump.
This is the sign that is most often got wrong, so it is a choice rather than a minus key. Signed value in use: -8 ft.
A property of the pump, not of your pipework, so there is no default for it here. Read it off the NPSHr line at the bottom of the manufacturer's curve at the flow above.
Bore 3.068 in, ASME B36.10M. Roughness 0.0457 mm.
Oversizing here is nearly always right: every foot of friction comes straight off the margin.
K 0.5 costs 0 ft at 0 ft/s. A bell mouth is nearly free and a pipe stuck through a tank wall is the worst of the three.
A foot valve and strainer is usually the biggest single resistance on a lift suction and this site does not carry a figure for one: the table below stops at the swing check. Take the equivalent length or the Cv from the valve maker's own data and add it here rather than letting a page guess it.
Fittings between the liquid and the pump
Crane rates these in pipe diameters, so in 3 in pipe one long-radius elbow is 5.11 ft of equivalent pipe. On a suction line the count is usually small and every one of them matters.
Blank uses the common rule of thumb — the greater of 35% of NPSHr and 3 ft, which for your 10 ft is — ft. ANSI/HI 9.6.1 is the actual standard and sets the margin by suction energy and service, from about 1.1 for a low-energy intermittent duty to 2.0 and beyond for high-energy or critical service.
Net positive suction head available
NPSH AVAILABLE
—
—
| Term | Feet | Meters | Reading |
|---|
Water density, viscosity and vapor pressure interpolated from a 0–100 °C anchor table — density linearly, viscosity and vapor pressure on their logarithms, which holds the mid-interval vapor error near 0.2%. Barometric pressure from US Standard Atmosphere 1976, troposphere: p = 101.325 kPa × (1 − 2.25577 × 10⁻⁵ h)^5.25588 with h in meters. Gives 14.70 psia at sea level, 12.23 psia at 5000 ft and 10.11 psia at 10 000 ft, matching the published tables. It is the standard-day atmosphere, not today's weather — a deep low runs about 0.5 psi below it, which is another foot of head off the suction side. Suction friction by Darcy-Weisbach over 36.3 ft of equivalent Steel, Schedule 40 3 in, plus the entrance at K 0.5. Pump suction line guidance is 5 ft/s (2–5 ft/s): Lower than the discharge limit on purpose. Friction head on the suction side subtracts directly from NPSH available, and a suction line is the one place where oversizing the pipe is nearly always right.
The same system, from room temperature to the boil
Nothing changes down this table except the liquid's temperature. The vapor term rises faster than anything else on the page and takes the margin with it, which is why a transfer pump that has run for years on cold water starts hammering the week it is put on a hot loop.
| Temperature°F · °C | Vapor pressurepsia · kPa | Vapor headft · m | NPSH availableft · m | Margin over NPSHrft |
|---|---|---|---|---|
| Complete the suction side above and this table fills in. | ||||
Velocity in the suction line is 0 ft/s · 0 m/s, and the flow regime behind the friction factor is on the Reynolds number calculator. For a domestic well, where the pump is under the water rather than above it, the whole suction question turns into a submergence question on the well pump size calculator.
The atmosphere can hold a column of cold water about 34 ft high at sea level, and that is the absolute ceiling on suction lift for any pump of any size — a bigger pump does not lift water further, because it is the air pushing, not the pump pulling. Subtract the vapor-pressure head, the suction friction and the pump's own NPSH required and the practical limit for a surface pump on cold water is nearer 20–25 ft, less at altitude and much less on hot water. Below that, the pump goes down the well rather than the water coming up to it, which is why deep wells use submersibles.
How to check NPSH available against a pump
Three things decide it: what pushes the liquid in, what the liquid does about it, and what the pipe takes back.
State the liquid and where it sits
Temperature first, because the vapor term moves faster with it than anything else on the page. Then the elevation of the installation for the barometric term, whether the vessel is open, pressurized or under vacuum, and whether the liquid surface is above the pump centerline or below it. That last one is a selector rather than a minus sign, because a sign error here is the single most common way this calculation goes wrong and it produces a plausible answer.
Build the suction line
Pick the pipe family and size, give the developed length, choose how the liquid enters the pipe and count the fittings. Add anything the fitting table does not carry — a foot valve and strainer above all — as equivalent length by hand. Every foot of friction on this side comes straight off the margin, which is why suction pipe is routinely a size larger than the discharge.
Enter NPSH required and read the margin
NPSH required is a property of the pump, so it comes off the curve, and it is read at the highest flow the pump will ever see rather than at the design point — it climbs with flow while the available falls with the square of it. The page then reports the margin in feet, the available-to-required ratio, and the greatest height the pump may sit above the liquid and still hold the margin you set.
Technical specifications
| The equation | NPSHa = (p_surface,abs − p_vapor) ÷ γ + z − h_f, with every term converted to feet of the liquid being pumped and reported on its own line rather than folded into the total. |
|---|---|
| Water properties | Density, dynamic viscosity and saturation pressure interpolated from 21 anchors at 5 °C steps between 0 and 100 °C. Density interpolates linearly; viscosity and vapor pressure interpolate on their logarithms, which holds the mid-interval vapor error near 0.2% instead of about 0.9%. |
| Vapor pressure span | 0.339 psia at 68 °F rising to 14.696 psia at 212 °F — a factor of 43 across the range building and process water actually sees. At the boil it cancels the atmospheric term exactly and no suction lift of any kind is possible. |
| Barometric pressure | US Standard Atmosphere 1976 troposphere formula: 14.70 psia at sea level, 12.23 psia at 5,000 ft, 10.11 psia at 10,000 ft. About 1 psi per 2,000 ft, which is 2.3 ft of head gone before the suction pipe has been counted. |
| Entrance coefficients | Crane TP-410 fixed-K values: 0.04 for a well-rounded bell mouth, 0.5 for a sharp-edged entrance, 0.78 for a pipe projecting inward. Applied to the actual velocity in the bore, not to a nominal one. |
| Margin rule | Prefilled as the greater of 1.35 × NPSHr and 3 ft, and editable. ANSI/HI 9.6.1 is the standard that governs it and sets margins by suction energy and service, from roughly 1.1 on a low-energy intermittent duty to 2.0 and beyond on high-energy or critical service. |
| What this page will not supply | A foot valve or strainer resistance. The fitting table stops at a fully open swing check at 100 pipe diameters, and a foot valve is usually the largest single loss on a lift suction — take its equivalent length or Cv from the valve maker and enter it by hand rather than accepting a guessed one. |
| Where the arithmetic runs | Entirely in this browser tab. Temperatures, elevations, pipe runs and the NPSHr figure you read off a curve are never transmitted, so the page answers just as well from a pit with no reception. |
Frequently asked questions
Is the static head positive or negative on a suction lift?
Negative — it is subtracting from what the atmosphere gives you, so a pump 8 ft above the water surface starts 8 ft down. The sign convention that makes this trip people is that z is measured from the liquid surface to the pump centerline and is positive when the liquid is the higher of the two, which is a flooded suction. This page asks with a selector rather than a minus key for that reason: the wrong sign here shifts the answer by twice the lift and the result still looks like a reasonable number.
Why does the same pump cavitate on hot water when it was fine on cold?
Because the liquid's vapor pressure is subtracted from what pushes it into the pump, and that term rises roughly exponentially with temperature. At 68 °F it costs about 0.8 ft of head; at 140 °F about 6.8 ft; at 212 °F about 35 ft, which is the entire atmospheric term. Nothing else in the installation has changed — same pipe, same lift, same pump — and the margin has simply been eaten. The temperature sweep on this page exists to show where in that range a given system runs out.
Which flow do I read NPSH required at?
The highest flow the pump will ever run at, not the design point. NPSH required is a curve, not a number: it rises with flow, because the faster liquid accelerates into the impeller eye the further its local pressure drops. NPSH available moves the opposite way, falling with the square of flow as suction friction climbs. The two worst cases therefore coincide, and a pump checked only at its design flow can cavitate every time a valve opens somewhere else on the system.
How high above the water can a pump physically sit?
About 34 ft of cold water at sea level, and that is the ceiling for any pump ever built, because it is the atmosphere doing the pushing rather than the pump doing any pulling. Subtract the vapor head, the suction line's friction, the pump's own NPSH required and a working margin, and the practical figure for a surface pump on cold water is nearer 20 to 25 ft. It falls with altitude, and it falls off a cliff with temperature. Below that limit the pump goes to the water instead, which is what a submersible is.
Would a bigger pump lift the water further?
No, and this is the most durable misconception in pumping. Suction lift is set by the pressure difference across the liquid column and nothing else, so a 50 hp pump lifts water from exactly the same depth as a 1/2 hp pump. What a bigger pump buys is flow and discharge head, both on the other side of the impeller. If the lift is the problem, the answers are to lower the pump, cool the liquid, shorten or enlarge the suction pipe, or pick a pump with a lower NPSH required — never a larger motor.
How much margin over NPSH required is actually enough?
It depends on suction energy, duty cycle and what failure costs, which is why ANSI/HI 9.6.1 exists rather than a single number. The rule of thumb this page prefills — the greater of 35% and 3 ft — is a defensible starting point for ordinary continuous water pumping and nothing more. Note also what NPSH required means: it is the point at which the pump has already lost 3% of its head to cavitation, so running exactly at NPSHr is running a pump that is quietly cavitating, not one on the edge of it.
Does altitude matter if the pump is flooded?
Yes, on any vessel open to the atmosphere. The barometric term is the biggest single contributor to NPSH available and it falls about 1 psi per 2,000 ft of elevation, which is 2.3 ft of water head. A flooded suction with 3 ft of margin at sea level has none of it left at 5,000 ft: the barometric term alone is 5.7 ft lower there, and nothing else in the installation has changed. The exception is a sealed pressurized vessel, where the absolute pressure on the surface is set by the vessel rather than by the weather — though a vessel holding liquid at its own boiling point gains nothing from either.
About net positive suction head, and why it is a system property
The two halves of this calculation belong to different owners, and confusing them is what produces most bad pump selections. NPSH available is a property of the installation: the barometer, the temperature of the liquid, the height of the surface relative to the pump and the pipe in between. Nothing about the pump appears in it. NPSH required is a property of the pump alone, measured on a test rig and drawn on the manufacturer's curve, and it is defined as the suction head at which the pump has already lost 3% of its head to cavitation. That definition matters more than it sounds: a pump run at exactly its NPSHr is not on the threshold of cavitating, it is cavitating, and the margin between the two curves is what buys quiet running and bearing life rather than a safety factor against nothing.
The vapor term is where the physics stops being intuitive. Cavitation is not air being sucked in; it is the liquid itself boiling at ambient temperature because the local pressure at the impeller eye has fallen below its saturation pressure, and then collapsing violently a few millimeters later where the pressure recovers. That is why the damage is pitting on the impeller vane rather than anything at the suction flange, and why the noise is described as gravel. It also explains the hardest case on this page: a liquid that is already at its boiling point in the vessel — a deaerator, a condenser hotwell, a flash tank — has its pressure term and its vapor term cancel exactly, so every foot of NPSH available has to be bought with static column, which is why those vessels are mounted high up in the plant and their pumps are in the basement.
On the suction side, Hydraulic Institute practice is to size generously and keep it short: a suction line one nominal size larger than the discharge, a long-radius elbow rather than a standard one, an eccentric reducer at the pump with the flat side up so no air pocket can form, and no high point anywhere in the run. The reason is the arithmetic above — friction here is subtracted from the margin, whereas friction on the discharge only costs pumping energy, and the two sides of the same pump are therefore designed to different rules. What the pump has to make once the liquid is inside it is the pump head calculator; the flow regime the friction factor came from is on the Reynolds number calculator, and the volume of liquid the suction line itself is holding is on the pipe volume calculator. A domestic well sidesteps the whole question by putting the pump under the water, and what replaces it there — submergence, drawdown and the drop pipe — is on the well pump size calculator. None of this is a stamped design: the numbers here are tables and arithmetic, and the pump manufacturer's curve and a qualified engineer decide the selection.
Where the suction figures are worked out
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Nothing you type — loads, lengths, nameplate ratings, the rates your utility charges you — is uploaded, logged or kept, which is also why the calculators carry on working in a mechanical room with no signal.