NPSH Calculator: Available Head and Cavitation

NPSH available is absolute pressure at the suction surface, minus the vapour pressure of the liquid, both converted to feet of head, plus the static height of liquid above the pump centreline, minus the friction losses in the suction line. Everything is expressed in feet of the fluid being pumped. It is a property of the system, not of the pump.

NPSHa is what the system provides and NPSHr is what the pump demands, published by the manufacturer as the point where head has already fallen three percent from cavitation. NPSHa must exceed NPSHr with a real margin, commonly a few feet or a ratio above about 1.2. Meeting NPSHr exactly means accepting a pump already three percent into cavitation.

Because vapour pressure climbs steeply with temperature and it is subtracted directly from available head. Water at 20°C has a vapour pressure of 0.34 psia and leaves about 40 feet available in a typical flooded suction. The same arrangement at 80°C has a vapour pressure of 6.88 psia and only about 26 feet. At 95°C it falls to roughly 13 feet. Nothing about the piping changed; the fluid did.

Because hot WFI distribution is deliberately held around 80°C or above for microbial control, which is precisely where vapour pressure has eaten a third or more of the available suction head. A distribution pump that behaves perfectly on a cold water flush can cavitate as soon as the loop comes up to temperature, and the fault appears to be intermittent when it is simply temperature dependent.

Vapour bubbles form where the local pressure falls below vapour pressure, then collapse violently as pressure recovers in the impeller. The collapses pit the impeller and casing, produce a distinctive gravel-in-the-pump noise, and erode the surface finish. On a hygienic pump that last point matters beyond mechanical wear, because a pitted product-contact surface is no longer a cleanable surface.

Attack the terms in the equation. Raise the liquid level or lower the pump to add static head. Shorten, straighten and enlarge the suction line to cut friction losses. Cool the fluid if the process allows it, which reduces vapour pressure sharply. Reduce flow, since NPSHr rises with capacity. Throttling the suction is the one move that always makes it worse.

Yes, and it is often missed because the pump was specified at a different site. Atmospheric pressure falls roughly half a psi per thousand feet of elevation, and that comes straight off NPSH available. A suction arrangement with two feet of margin at sea level can be marginal in Denver, and a design that was fine at one plant can fail when it is copied to another.

Only as a method, not as numbers. The table is water. Every term must be evaluated for the actual fluid: vapour pressure at the operating temperature, density for the head conversion, and viscosity for the friction losses. Solvents and concentrated process solutions can have vapour pressures far above water at the same temperature, which is why they cavitate at conditions that would be comfortable for a water system.

A dimensionless index built from flow, NPSH required and pump speed that indicates how hard the impeller eye is working. High values mean an impeller designed for low NPSH requirement, which usually narrows the flow range over which the pump runs quietly. It is a useful warning that a pump with an unusually low published NPSHr may be intolerant of off-design operation.

The lowest rate at which a pump can run without internal recirculation and overheating. Throttling a cavitating pump down to reduce NPSH demand can push it below this limit, replacing one damage mechanism with another. The correct fix is usually a smaller impeller or a variable speed drive, not a closed discharge valve.

Yes, and it is often more demanding than a centrifugal at the same duty because the inlet must fill a chamber within a fixed time. Lobe and diaphragm pumps used on hygienic duty are frequently limited by suction conditions rather than by discharge capability, which is why their suction lines are short and generously sized.

More than zero, because NPSHr is defined at three percent head loss and the pump is already cavitating at that point. Common practice is a margin of several feet, or a ratio of available to required above roughly 1.2, with more on high-energy pumps and on any duty where the fluid is near its boiling point.

Put a gauge at the pump suction flange and read it with the pump running at duty. Convert the absolute pressure to head in the actual fluid, subtract the vapour pressure head at the measured temperature, and add the velocity head. That gives the real NPSHa including whatever the suction piping is actually doing, rather than what the drawing assumed.

Yes, and it is one of the most common causes of a pump that was fine last month. A fouling strainer adds friction loss directly to the suction side, subtracting from NPSHa. The symptom develops gradually and is often mistaken for pump wear, which is why suction strainer differential is worth trending on critical duty.

It does. An elbow immediately upstream delivers a distorted, swirling profile into the impeller eye, which behaves like a local reduction in available NPSH and can also produce unbalanced radial loads. Several diameters of straight pipe before the flange is standard practice, and on double-suction pumps an elbow in the wrong plane is particularly damaging.

To avoid trapping air. A concentric reducer on a horizontal suction line creates a high point at the top where vapour collects and eventually enters the impeller. An eccentric reducer fitted flat side up keeps the top of the line continuous so nothing can pocket, which is why it is specified almost universally on horizontal suctions.

Cavitation is the fluid boiling and re-condensing because of low local pressure. Air entrainment is free gas being drawn in from a vortex, a leaking joint or a tank level that has fallen too far. They sound similar and both damage performance, but the remedies differ entirely: one is a pressure problem, the other is a sealing or level problem.

About 33 feet at sea level for cold water, since that is the height of a water column atmospheric pressure can support. Practical lift is well below that once vapour pressure, friction and the pump’s own requirement are subtracted, and it falls sharply with temperature. Hot water effectively cannot be lifted at all, which is why hot systems use flooded suctions.

Mainly through friction. A viscous fluid loses far more head in the suction line for the same velocity, which reduces NPSHa. Viscosity also alters pump performance, derating head and capacity, so a pump selected on water curves can fail on both counts when the real fluid is thick.

Often yes, because NPSH required falls as speed drops. Running slower reduces both the demand and the flow-dependent friction loss in the suction line. It is not a cure for a fundamentally bad suction arrangement, and the minimum continuous flow limit still applies at the reduced speed.

Like gravel or marbles passing through the pump, produced by the implosion of vapour bubbles rather than by any mechanical contact. It is usually accompanied by unsteady discharge pressure and vibration. The sound is diagnostic enough that an experienced operator can identify it before any instrument shows a problem.

By destroying the surface finish. The implosions pit the impeller and casing, and on a product-contact surface a pitted finish is no longer a cleanable one. A cavitation problem on a sanitary pump therefore escalates from a mechanical issue to a cleaning-validation issue, and the component usually has to be replaced rather than dressed.

Yes, favourably. NPSHa uses the absolute pressure above the liquid, so a blanketed or pressurised vessel adds directly to available head. This is why nitrogen-blanketed tanks often solve suction problems that no amount of piping change would fix, and why losing blanket pressure can make a previously reliable pump cavitate.

Because the velocity into the impeller eye rises with capacity, and higher velocity means a lower local pressure at the point where the fluid turns into the vanes. A pump run far out on its curve therefore demands more suction head at exactly the moment the suction line friction is also at its highest.

On a gravity transfer, rarely, because there is no impeller creating a low-pressure region. It becomes a concern the moment a pump is introduced, particularly on hot WFI where vapour pressure has already consumed most of the margin. Gravity transfer of hot liquid into a pump suction is a classic arrangement for marginal NPSH.