Source

LIQUID LEVEL ABOVE PUMP CENTRELINE TEMPERATURE PRESSURE ON THE LIQUID SURFACE

Suction line

LINE SIZE DEVELOPED LENGTH ELBOWS + VALVE + STRAINER

Duty

FLOW PUMP NPSHr AT RATED FLOW (200 m³/h)

Reading it

Model: NPSHa = (P_surface − P_vapour)/(ρg) + h_static − h_friction. Water properties from Antoine (Stull) and a saturated-water density/viscosity table; friction by Darcy–Weisbach with Swamee–Jain f, ε = 0.046 mm, each fitting at 30 D of equivalent length, plus an entrance/reducer allowance of K = 0.6. Pump demand scaled as NPSHr ∝ Q^1.8 about the rated point — a typical centrifugal shape, not a real curve. Velocity head at the flange is deliberately not added: NPSHa is referenced to the pump suction flange and includes it on both sides of the comparison.

The reducer is not in this arithmetic — and that is the point. A concentric reducer on a horizontal suction costs you almost nothing in head loss, so it never shows up in an NPSH calculation. It fails you a different way: it builds a high point that collects the vapour this calculation says you are already close to releasing, and sheds it into the eye in slugs. Margin protects against the average; the reducer decides what the impeller sees instant by instant.