Suction lift
Size a self-priming pump for suction lift
Last reviewed 21 October 2026.
Suction lift is the ordinary layout for a self-priming pump above a wet well, a sump, or a tanker. The pump sits where you can service it. The liquid sits below. The static head includes every foot of that difference, and the suction line has to be emptied of air before the pump can act like the curve you were shown. Sizing that duty is a system-curve problem and an NPSH problem at the same time. Ignore either one and the pump that “made the flow on the bench” will not make it in the hole.
This guide is about the decisions that belong in that sizing, using CurvePoint’s definitions: pump suction centerline as the datum, Darcy-Weisbach friction, and NPSH from atmosphere, vapour, and total dynamic suction lift. The numbers you should calculate live on the system curve calculator and the NPSH calculator. A finished fictional file, 4 in steel at a 12 ft lift, is the wastewater example.
Design at the lowest liquid level
| Level | Static | Operating point | NPSH available |
|---|---|---|---|
| Low water, 16 ft lift | 38 ft (11.58 m) | 509 gpm (32.1 L/s) at 64.0 ft (19.49 m) | 15.4 ft (4.70 m) |
| High water, 6 ft lift | 28 ft (8.53 m) | 549 gpm (34.6 L/s) at 58.1 ft (17.71 m) | 25.3 ft (7.72 m) |
Illustrative Lakeshore Process Equipment figures from CurvePoint’s own evaluation. Not a sealed calculation.
The table is one Lakeshore line at two well levels. Low water is a 16 ft lift. High water is a 6 ft lift. Flow, head, and available NPSH all move. The low-water row is the one to design. The high-water row is what the same pump does when the well has recovered. A storm allowance belongs in that low-water flow: Ontario’s water-taking guide asks for the taking you will actually make, including storm water that shows up whether or not you planned on it.
Self-priming is a pump feature, not a free lift
A standard end-suction pump that is full of liquid will develop the head on its curve. It will not, by itself, evacuate a dry suction line. If the casing is opened to atmosphere and the suction pipe is full of air, the impeller is moving gas, and the head it can produce in feet of gas is useless against feet of liquid. A self-priming centrifugal pump is arranged so that it can retain a charge of liquid, separate the air it pulls from the suction line, and send that air out the discharge until liquid follows. After that, it runs as a centrifugal pump on the system curve. The priming device does not add head to the curve, and it does not cancel suction lift.
Reprime matters as much as the first prime. A lift station that stops with the suction line full will start easily. One that drains back, or that swallows a slug of air when the wet well level bounces, has to prime again under the same lift. A foot valve or a flap that holds the column can make reprime faster, and it also adds loss and a maintenance item. If the valve leaks, you have paid the K and you still reprime. Put that loss in the suction friction when you claim the benefit. CurvePoint’s self-priming pump type keeps a suction line in the study for exactly this layout. The submersible type removes it, because the pump is already in the liquid.
The lift you can use is much less than the theoretical ceiling
Cold water at sea level has an atmospheric head a little above 33 ft. That figure is a ceiling, not a design lift. Vapour pressure takes a slice. Suction friction takes another, and it grows as soon as you ask for flow. Altitude lowers the starting atmospheric head before you draw the well. Temperature raises vapour pressure. By the time those terms are honest, a static lift in the mid-teens is already a serious duty, and a static lift near 25 ft is a duty you should try hard not to build. Many wastewater specifications keep static lift under 20 ft so the pump can reprime and still show NPSH margin at the high-flow end of the curve.
Measure lift from the pump centerline to the lowest liquid surface you must pump, not to the invert of the inlet and not to the average level. The worst NPSH and the longest prime both happen at low water. If the well is deep and the pump can be lowered, lowering it is often cheaper than buying a curve with more NPSH margin. If the pump must stay at grade, the suction line has to be designed as if friction were expensive, because it is. Use the NPSH calculator with your atmospheric pressure, that low-water lift, a realistic suction friction, and the vapour pressure at the hottest fluid you expect. If the estimate is thin before you have even read the curve, change the layout.
Size the suction line for velocity, then check the system curve
A suction line that matches the discharge nozzle is a common and often poor default. The discharge can afford velocity. The suction cannot. A larger suction diameter cuts velocity, cuts friction, and cuts the chance that a small air pocket will separate and break prime. Keep the run short. Use long-radius bends instead of a stack of standard elbows at the pump. Avoid tees, reducers that are abrupt, and anything that can trap air at a high point. Slope the pipe up toward the pump so air moves to the casing instead of sitting in a pocket. A high point between the well and the pump is a prime killer.
Then put that line on a system curve. Static head is lift plus discharge elevation. Friction is suction plus discharge, and the discharge on a force main is often where the length actually lives. A 15 ft suction and a 250 ft discharge, both 4 in steel, is the shape of the sample study: the suction loss at a few hundred gallons per minute is about a foot and a half, and the discharge loss is an order of magnitude more. The suction loss is the one that hits NPSH. The discharge loss is the one that dominates total head. You need both, and they are not interchangeable. Build them in the system curve calculator before you fall in love with a shutoff head.
Air, solids, and gas change a lift that looked fine on water
An air leak on the suction side of a flooded pump is a nuisance. On a lift it is the failure. Clamps, threaded joints, a worn stuffing box, and a suction gauge cock left cracked will admit air because the pressure inside the pipe is below atmosphere. The pump may still move some liquid while it swallows air, and the curve you plotted no longer describes the head. Pressure-test the suction line. Do not rely on a successful prime once, on a new gasket, as proof the line will stay tight.
Wastewater adds ragging, grit, and gas that comes out of solution when the pressure drops in the suction line. Dilute industrial wastewater, under about 2 percent solids, is treated as hydraulically close to water in CurvePoint, with a warning rather than a silent derate of the pump curve. Heavier sludge is a different friction model and often should not be on a lift at all. If the liquid foams or if the well is septic, the vapour-pressure term you used for clean water is optimistic. Raise it, or stop claiming the clean-water NPSH. The sample report is explicit that it is an estimate for a fictional dilute wastewater at 59°F, not a permit to lift thicker stock the same way.
A sizing sequence that survives review
Write the low-water lift and the discharge static head first, in feet from the pump centerline. Pick a suction diameter that keeps velocity modest, sketch every fitting, and estimate suction friction at the highest flow you will accept, not only at the design point. Estimate discharge friction on the real force main, including the outlet. Add static and friction into a system curve. Only then open pump curves, at the speed you will run, and look for an intersection inside the preferred range. Read NPSH on the curve at that intersection and compare it with available NPSH at the same flow. If you must throttle or slow the pump to hit a downstream limit, move the intersection on purpose, as the operating point guide describes, and recheck NPSH after the move.
CurvePoint is built for that sequence. Choose a self-priming pump, enter the lift and both lines, and plot or import the curve. The study on screen is free. The PDF report, with your name on it, is the Pro export described on pricing. Start from the homepage sample if you want to see a lift station written out before you enter your own elevations, or open the calculator and begin with the water and wastewater suction-lift preset. Either way, mark the results as estimates until someone has checked them against the manufacturer’s curve and the pipe that was actually installed.
What still has to be true in the hole
Suction velocity stays modest. A 6 in hose that is asked to carry the discharge duty of a much larger pump will show it in the friction term and in the tendency to trip air. Air leaks matter on a lift because the pipe is below atmospheric pressure: joints, a gauge cock, and a worn seal admit air that the curve never included. A foot valve or a strainer can hold a column for reprime and can also add loss and a place for rags to sit. Count that loss, and clean the screen. Where the pump nozzle is smaller than the suction line, an eccentric reducer with the flat side up keeps a pocket of air from sitting at the crown. Those habits are the piping article’s, and they belong in the sketch before the submittal.
Where the water is dirty, the intake itself is part of the suction. DC’s dewatering strategy asks for a floating, screened intake. Ohio’s dewatering measures say to draw from near the surface so the pump does not pull fines off the sump floor. Both are submergence and water-quality choices. They show up in NPSH only after you have also kept the inlet covered.
Priming lift and NPSH answer different questions
Priming asks whether the pump can clear the hose of air and whether it can do that again after it loses prime. NPSH asks whether liquid that has already arrived will stay liquid at the impeller. A pump can prime and still run short of NPSH, and a flooded pump can have NPSH and still be unable to prime a dry hose. The budget below is only the atmospheric side: head left after vapour pressure, before you spend any of it on lift, hose friction, or the pump’s required NPSH.
| Elevation | Water at 10°C | Water at 30°C |
|---|---|---|
| 0 m (0 ft) | 33.5 ft (10.21 m) | 32.6 ft (9.94 m) |
| 300 m (984 ft) | 32.3 ft (9.85 m) | 31.4 ft (9.58 m) |
| 600 m (1969 ft) | 31.2 ft (9.50 m) | 30.3 ft (9.23 m) |
| 1000 m (3281 ft) | 29.7 ft (9.04 m) | 28.8 ft (8.77 m) |
| 1500 m (4921 ft) | 27.9 ft (8.50 m) | 27.0 ft (8.23 m) |
Illustrative Lakeshore Process Equipment figures from CurvePoint’s own evaluation. Not a sealed calculation.
Three machines clear that hose in different ways. A wet-prime self-primer keeps a charge of liquid in the casing and uses it to sweep air out the discharge. A prime-assisted dewatering pump evacuates the casing and the hose with a compressor and venturi, or with a diaphragm vacuum pump, and a float valve should close once liquid arrives. The dewatering guide is that layout, including how elevation spends the same budget. A vane-assisted primer on a tanker offload is a small positive-displacement air mover on the same shaft, used until liquid reaches the impeller. The tanker guide sizes the hose that primer has to empty. None of the three adds head to the centrifugal curve.