Independent study aid. Not affiliated with or endorsed by NCEES. Always verify against the current NCEES exam specifications and reference handbook.
🎥 New: a free FE/PE worked-problems video every day — watch today’s video on YouTube

Theory explainer

NPSH Problems — FE/PE Civil Exam Worked Examples

A pump does not pull water up its suction pipe. The atmosphere pushes it, and the pump can only take what the atmosphere can deliver. Net positive suction head is the exam’s way of accounting for that push: how much pressure head the system delivers to the pump inlet, above the point where the liquid boils. Get the accounting wrong and the pump cavitates — bubbles form at the impeller eye, collapse violently, and eat the impeller while the head curve quietly collapses. Every NPSH problem below is the same four-term subtraction, and the exam’s traps are the same four terms: gauge instead of absolute pressure, the wrong sign on the suction lift, the vapour pressure forgotten, and the required NPSH read at the wrong flow.

Last reviewed: 2026-11-09. Every worked number below was independently re-solved; steps are shown in full.

What cavitation is, and why NPSH exists

Inside a centrifugal pump the pressure is lowest at the eye of the impeller, where the liquid accelerates into the blades. If that local pressure falls below the liquid’s vapour pressure at its temperature, the liquid boils without being heated: vapour bubbles form, are carried a short distance into the higher-pressure region of the impeller, and collapse violently. The collapse chips away at the impeller surface (pitting), makes a noise like gravel in the casing, vibrates the pump, and drops the head and flow the pump can deliver. That is cavitation, and it is a suction-side problem — nothing on the discharge side causes it. The exam’s defence against it is a single comparison: the suction head the system delivers, measured above the vapour pressure, must exceed the suction head the pump needs. The first quantity is NPSHA and you calculate it from the installation; the second is NPSHR and the manufacturer measures it on a test rig and prints it on the pump curve, where it climbs with flow. The comparison is done in head — feet or metres of the liquid being pumped — never in pressure, which is why every term in the NPSHA formula is a pressure divided by the specific weight γ.

NPSHA: the four terms, and the two tables that feed them

NPSHA = patm/γ ± hs − hL,suction − pv/γ. Each term is a head of the liquid being pumped, and each has its own trap. The atmospheric head is the push: 14.7 psi at sea level, which is 14.7 × 2.31 = 33.9 ft of water (101.3 kPa, or 10.3 m). It falls with altitude, and the exam’s favourite way of cutting NPSHA is to move the installation up a mountain — at 5,000 ft the atmospheric head is only 28.2 ft. The static suction head hs carries a sign, and the sign is the commonest arithmetic slip in the whole topic: if the source surface sits above the pump centreline (flooded suction) the head adds; if the pump sits above the source (suction lift) it subtracts. Draw the sketch before assigning the sign, exactly as the Pumps & System Curves page advises. The suction loss hL,suction is the friction and minor loss in the suction piping only — not the whole system — and it always subtracts, because energy spent fighting pipe wall and fittings on the way to the impeller is energy no longer available as pressure. The vapour-pressure head pv/γ is the term candidates forget, and it is the one that does the damage: it is read from the steam tables at the pumping temperature, it must be absolute, and it rises steeply with temperature.

Water temperatureVapour pressureVapour-pressure head
32 °F (0 °C)0.089 psi0.21 ft
60 °F (16 °C)0.256 psi0.59 ft
68 °F (20 °C)0.338 psi0.78 ft
100 °F (38 °C)0.949 psi2.19 ft
150 °F (66 °C)4.36 psi10.07 ft
200 °F (93 °C)11.53 psi26.63 ft
212 °F (100 °C)14.70 psi33.95 ft

Read the table as a warning: the vapour-pressure head is under a foot for cold water and over ten feet at 150 °F. The suction pipework does not change at all across that range — the temperature alone eats ten feet of margin, which is why Worked Problem 3 cavitates and why any exam stem that gives a water temperature is telling you to look up the vapour pressure. The altitude table is the same story on the atmospheric side:

Elevation above sea levelAtmospheric pressureAtmospheric head
Sea level14.70 psi33.9 ft
2,000 ft13.66 psi31.6 ft
5,000 ft12.23 psi28.2 ft
10,000 ft10.11 psi23.4 ft

One conversion to keep straight: 1 psi = 2.31 ft of water, because 1 psi × 144 in²/ft² ÷ 62.4 pcf = 2.31 ft. That factor is for water at 62.4 pcf; for another liquid, divide by its specific gravity instead — a liquid with specific gravity 1.2 has an atmospheric head of 33.9 ÷ 1.2 = 28.2 ft, the same number as the 5,000 ft altitude case, by coincidence and by the same arithmetic. In SI units the same conversion is a pressure divided by ρg: 101.3 kPa ÷ 9.81 kN/m³ = 10.3 m of water.

NPSHR: a pump property that climbs with flow and speed

NPSHR is not calculated from the system — it is read from the pump curve at the operating flow, where it is usually plotted as a separate, dashed line that climbs across the flow range. It is defined by the Hydraulic Institute / API 610 convention as the suction head at which the pump’s developed head has already dropped by 3% through cavitation onset (NPSH3), which is the single most important fact about it: at exactly NPSHA = NPSHR the pump is already cavitating, mildly. NPSHR is the threshold of measurable damage, not the safe operating limit, which is why the design rule is NPSHA > NPSHR with a margin of 2–5 ft (0.6–1.5 m), and why the Hydraulic Institute recommends at least 0.6 m and API 610 the larger of 1 m or 1.1 × NPSHR. Two things move NPSHR. First, flow: NPSHR rises with flow, because more flow means higher velocity at the impeller eye and more of the suction pressure converted to kinetic energy — running a pump far to the right of its best-efficiency point, towards runout, is a cavitation risk even in a system that was safe at the duty point. Second, speed: under the affinity laws NPSHR scales with the square of the speed, NPSHR2 = NPSHR1 × (N2/N1)², exactly like head. Worked Problem 4 is built on that scaling: a speed increase that looks like free capacity quietly raises the required NPSH past what the system delivers. Note the asymmetry — NPSHA does not change with speed, except that the higher flow raises the suction friction loss and trims NPSHA a little further, so a speed increase attacks the margin from both sides at once.

The maximum suction lift, and what to do when the margin fails

Set NPSHA = NPSHR and solve the four-term equation for the lift, and you get the maximum suction lift a pump can tolerate: hs,max = patm/γ − pv/γ − hL,suction − NPSHR. At sea level with cold water that is roughly 33.9 − 0.6 − a few feet of loss − the required NPSH — and in practice a surface pump rarely lifts cold water more than about 25 ft (7–8 m), because the losses and the required NPSH eat the rest. The theoretical ceiling is the atmospheric head itself: no pump can lift water by suction more than 33.9 ft at sea level, because beyond that the water vaporises in the suction pipe and the pump loses its prime. That ceiling is why deep wells use submersible pumps, placed down in the water, rather than surface pumps — see the Groundwater Flow page for the well hydraulics. When a problem shows NPSHA below NPSHR, the fixes follow the four terms: lower the pump (turn a suction lift into a flooded suction — the biggest single gain, a foot for a foot), cool the liquid (cut the vapour-pressure head), enlarge the suction pipe (cut the friction loss), slow the pump down or throttle the flow (cut NPSHR), or pressurise the source tank (raise the absolute pressure on the surface). Worked Problem 3 ends with the first of these: the maximum lift that would work at that altitude and temperature is 1.18 ft, so the pump as installed — 6 ft above the source — has to move down, or the water has to cool.

Worked problem 1 Flooded suction: NPSHA and the margin (FE pace, US units)

Given: a centrifugal pump draws water at 60 °F from an open tank whose surface sits 4 ft above the pump centreline. The suction line loses 1.5 ft of head. The pump curve gives NPSHR = 12 ft at the operating flow. The installation is at sea level. Find NPSHA and decide whether the pump cavitates.

Solution:

  1. Atmospheric head at sea level: patm/γ = 14.7 psi × 2.31 ft/psi = 33.9 ft.
  2. Vapour-pressure head at 60 °F: pv = 0.256 psi, so pv/γ = 0.256 × 2.31 = 0.59 ft.
  3. Static suction head: the source is above the pump, so this is a flooded suction and the sign is positive: hs = +4.0 ft.
  4. NPSHA = 33.9 + 4.0 − 1.5 − 0.59 = 35.81 ft.
  5. Compare: NPSHA = 35.81 ft against NPSHR = 12 ft. Margin = 35.81 − 12 = 23.8 ft — far above the 2–5 ft guideline.

Answer: NPSHA ≈ 35.8 ft; margin ≈ 23.8 ft over NPSHR = 12 ft — no cavitation. Trap note: this is the easy case, and it is easy precisely because every term helps — flooded suction adds, cold water costs almost nothing in vapour pressure, and sea-level air pushes hardest. The exam’s version of this problem changes one of those four things (a suction lift, hot water, altitude) and the answer swings by ten feet. The second trap is the gauge trap: if a stem gives a suction pressure as a gauge reading, add 14.7 psi before converting to head — NPSH is always absolute.

Worked problem 2 Suction lift: the sign that costs ten feet (FE pace, US units)

Given: the same pump is moved so its centreline sits 10 ft above the surface of an open tank of water at 80 °F. The suction line loses 2.5 ft of head. NPSHR = 15 ft at the operating flow, at sea level. Find NPSHA and the margin.

Solution:

  1. Atmospheric head: 33.9 ft, as in Worked Problem 1.
  2. Vapour-pressure head at 80 °F: pv = 0.507 psi, so pv/γ = 0.507 × 2.31 = 1.17 ft.
  3. Static suction head: the pump is above the source, so this is a suction lift and the sign is negative: hs = −10.0 ft.
  4. NPSHA = 33.9 − 10.0 − 2.5 − 1.17 = 20.23 ft.
  5. Margin = 20.23 − 15 = 5.23 ft — above the 2–5 ft guideline, just.

Answer: NPSHA ≈ 20.2 ft; margin ≈ 5.2 ft over NPSHR = 15 ft — no cavitation, but the margin is thin. Trap note: compare with Worked Problem 1 — the only structural change is the sign on hs, and NPSHA falls by 15.6 ft (10 ft of lift plus 0.58 ft of extra vapour pressure plus 1 ft of extra loss). The sign on the suction lift is the single most-tested NPSH trap: a candidate who adds the 10 ft instead of subtracting it gets 40.2 ft and a comfortable-looking margin that does not exist. The second trap is the margin judgement: 5.2 ft passes, but only just — an exam stem that then raises the flow, and with it NPSHR, flips this installation into cavitation, which is the bridge to Worked Problem 4.

Worked problem 3 Hot water at altitude: when the margin goes negative (PE pace, US units)

Given: a pump at 5,000 ft elevation draws water at 150 °F from an open tank, with its centreline 6 ft above the water surface. Suction-line loss is 3.0 ft. The pump curve gives NPSHR = 14 ft at the operating flow. (a) Find NPSHA and decide whether the pump cavitates. (b) What is the maximum suction lift that would work in this installation?

Solution:

  1. Atmospheric head at 5,000 ft: patm = 12.23 psi, so patm/γ = 12.23 × 2.31 = 28.25 ft — already 5.7 ft below the sea-level value before anything else happens.
  2. Vapour-pressure head at 150 °F: pv = 4.36 psi, so pv/γ = 4.36 × 2.31 = 10.07 ft.
  3. Static suction head: suction lift, so hs = −6.0 ft.
  4. (a) NPSHA = 28.25 − 6.0 − 3.0 − 10.07 = 9.18 ft. Against NPSHR = 14 ft the shortfall is 14 − 9.18 = 4.82 ft — the pump cavitates.
  5. (b) Set NPSHA = NPSHR and solve for the lift: hs,max = 28.25 − 3.0 − 10.07 − 14 = 1.18 ft. The pump would have to sit within 1.18 ft of the water surface — and that is with zero margin, so a safe installation needs it lower still, or the water cooler.

Answer: (a) NPSHA ≈ 9.2 ft, 4.8 ft short of NPSHR = 14 ft — the pump cavitates. (b) Maximum suction lift ≈ 1.2 ft at zero margin. Trap note: two terms do the damage here and neither is the pipework — the altitude costs 5.7 ft of atmospheric head and the temperature costs 10.1 ft of vapour-pressure head, together 15.8 ft before the lift and the friction are even counted. A candidate who uses the sea-level atmospheric head (33.9 ft) and forgets the vapour pressure gets 24.9 ft and a pass; the correct answer is a fail by nearly 5 ft. The second trap is part (b): the maximum lift is solved from the same four-term equation with NPSHA set equal to NPSHR — it is not a separate formula to memorise.

Worked problem 4 A speed increase moves the goalposts (PE pace, US units)

Given: a pump at 1,750 rpm has NPSHR = 9 ft at its duty point, and the system delivers NPSHA = 11 ft. The speed is raised to 2,000 rpm to get more flow. Does the pump cavitate at the new speed?

Solution:

  1. At 1,750 rpm the margin is 11 − 9 = 2.0 ft — safe, at the bottom of the 2–5 ft guideline.
  2. NPSHR scales with the square of the speed, like head: NPSHR2 = NPSHR1 × (N2/N1)² = 9 × (2,000/1,750)² = 9 × 1.3061 = 11.76 ft.
  3. NPSHA does not change with speed — the installation is the same — so at 2,000 rpm the comparison is 11 ft available against 11.76 ft required: a shortfall of 0.76 ft.

Answer: yes — at 2,000 rpm NPSHR ≈ 11.8 ft exceeds NPSHA = 11 ft by 0.8 ft, so the pump cavitates. Trap note: the speed increase is only 14%, but the required NPSH rises 31%, because the scaling is squared. The trap is treating NPSHR as a fixed pump property: it is fixed at a given speed and flow, and it moves with both. The exam’s favourite version of this problem gives NPSHR at one speed and asks for the maximum safe speed — solve NPSHR1 × (N2/N1)² = NPSHA for N2, which here is 1,750 × √(11/9) ≈ 1,934 rpm. Also note the second-order effect the stem does not ask for: the higher flow raises the suction friction loss, so NPSHA at the new speed is slightly below 11 ft and the real shortfall is worse than 0.76 ft.

Worked problem 5 NPSHA in SI units (FE pace, SI)

Given: a pump draws water at 40 °C from an open tank whose surface sits 2.0 m above the pump centreline. The suction line loses 0.6 m of head. The pump curve gives NPSHR = 4.5 m at the operating flow. The installation is at sea level. Find NPSHA and the margin.

Solution:

  1. Specific weight of water at 40 °C: ρ = 992 kg/m³, so γ = 992 × 9.81 = 9.73 kN/m³.
  2. Atmospheric head: patm/γ = 101.325 kPa ÷ 9.73 kN/m³ = 10.41 m.
  3. Vapour-pressure head at 40 °C: pv = 7.38 kPa, so pv/γ = 7.38 ÷ 9.73 = 0.76 m.
  4. Static suction head: flooded suction, so hs = +2.0 m.
  5. NPSHA = 10.41 + 2.0 − 0.6 − 0.76 = 11.05 m. Margin = 11.05 − 4.5 = 6.55 m.

Answer: NPSHA ≈ 11.1 m; margin ≈ 6.6 m over NPSHR = 4.5 m — no cavitation. Trap note: the SI version is the same four-term subtraction with pressures divided by ρg instead of multiplied by 2.31 — the arithmetic is identical, only the conversion changes. The trap is mixing the two: 101.3 kPa is not 10.3 m of water until it is divided by the specific weight, and the vapour pressure at 40 °C (7.38 kPa) is not negligible just because it looks small in kilopascals — it is 0.76 m of head, a sixth of the required NPSH. The second trap is the density: water at 40 °C is 992 kg/m³, not 1,000 — using 1,000 gives 10.33 m instead of 10.41 m, a small error here but the kind that compounds when the margin is thin.

Frequently asked questions

What is NPSH?

NPSH is net positive suction head: the pressure at a pump’s suction, expressed as a head of the liquid being pumped, above the liquid’s vapour pressure at its temperature. If the local pressure at the impeller eye falls below the vapour pressure, the liquid boils, bubbles form, and they collapse violently in the higher-pressure region — that is cavitation. NPSH comes in a pair: NPSHa (available) is what the system delivers, and NPSHr (required) is what the pump needs. The pump runs clean only when NPSHa exceeds NPSHr, with margin.

How do you calculate NPSHa?

NPSHa = Patm/gamma ± hs - hL,suction - Pv/gamma, all in feet (or metres) of the liquid. Patm/gamma is the atmospheric head (33.9 ft of water at sea level), hs is the static suction head — positive if the source sits above the pump (flooded suction), negative if the pump must lift (suction lift) — hL,suction is the head loss in the suction piping, and Pv/gamma is the vapour-pressure head at the pumping temperature. Every term must be in absolute pressure, and every term must be in the same head units. Worked Problem 1 runs the full calculation for a flooded suction.

What is the difference between NPSHa and NPSHr?

NPSHa is a property of the system: the source pressure, the height of the source relative to the pump, the suction pipework, and the liquid temperature. You calculate it. NPSHr is a property of the pump: the suction head the impeller needs at a given flow to avoid cavitation, measured on a test rig and read from the pump curve. It rises with flow and with the square of the speed. The design rule is NPSHa > NPSHr, with a margin of 2–5 ft (0.6–1.5 m) in practice — and at exactly NPSHa = NPSHr the pump is already cavitating mildly, because NPSHr is defined at the 3% head-drop point.

What is the maximum suction lift for a pump?

At sea level, pumping cold water, the theoretical maximum suction lift is the atmospheric head, 33.9 ft (10.3 m), minus the vapour-pressure head and the suction losses and the required NPSH. In practice a surface pump rarely lifts cold water more than about 25 ft (7–8 m), and hot water or high altitude cuts that sharply: at 150 degF the vapour-pressure head alone is 10.1 ft, and at 5,000 ft elevation the atmospheric head is only 28.2 ft. Worked Problem 3 shows a hot, high-altitude installation with only 9.18 ft of NPSHa against a 14 ft requirement — it cavitates — and the maximum lift that would work there is 1.18 ft.

Does NPSHr change with pump speed?

Yes — NPSHr scales with the square of the speed, like head, under the affinity laws: NPSHr2 = NPSHr1 x (N2/N1)^2. NPSHa does not change with speed (except that the higher flow raises the suction friction loss, which trims NPSHa a little further). So speeding a pump up can turn a safe installation into a cavitating one: Worked Problem 4 raises the speed from 1,750 to 2,000 rpm, NPSHr climbs from 9 ft to 11.8 ft, and an 11 ft NPSHa that was safe at the old speed is now short by 0.8 ft.

Why does hot water cavitate more easily than cold water?

Because the vapour pressure rises steeply with temperature, and the vapour-pressure head is subtracted from NPSHa. Water at 60 degF has a vapour-pressure head of 0.59 ft; at 100 degF it is 2.19 ft; at 150 degF it is 10.07 ft; at 200 degF it is 26.6 ft. The suction pipework does not change at all — the temperature alone can eat ten feet of margin. That is why boiler-feed and hot-condensate pumps run with far less margin than the same layout pumping cold water, and why exam stems that give a water temperature are telling you to look up the vapour pressure.

Free FE/PE flashcards + study tips, by email

Join the Water Resources Hub email list and get the free 40 water-resources flashcards (PDF + Anki) straight away, then study tips and practice problems by email. Email address only — no spam, unsubscribe any time.