Login

I'm trying to size a centrifugal pump for a closed-loop cooling system, but I'm getting conflicting results for the required net positive suction head. My calculations seem right based on the pipe friction losses and static head, yet the vendor's curve suggests my available NPSH is too low. I'm worried I'm missing something about the vapor pressure relationship with the system's temperature swing.

I’ve chased that same issue. The available NPSH isn’t only static head minus pipe losses; it shifts with the liquid temperature because vapor pressure climbs as it gets hotter. If the loop swings between cold and warm, you can lose margin fast even when friction looks okay. On a recent project we double checked Pv at the actual suction temperature and compared it to the vendor curve; we found the unit was flirting with cavitation at design speed. The fix wasn’t a magic new pump—we either cool the suction tank a bit or choose a pump with a larger NPSH margin and, if needed, add degassing to reduce air entrainment.

Another thing I spotted was measurement drift. We logged suction pressure while the pump ran and watched how the head losses changed with temperature; the NPSHA stayed low because the tank warmed during sidewall heat gain. We ended up cooling the tank or slowing the loop a notch to buy some margin rather than chasing a new curve.

Quick question: do you know if the vapor pressure curve you’re comparing against matches the actual coolant composition and temperature range in your loop? A mismatch there can make NPSH look worse than it is.

I had a moment where I thought the problem was the NPSH, but then I drifted to the idea that maybe the issue is air entrainment from a partially filled reservoir. Our tests showed some dissolved air coming out of solution as temperatures rose, which felt like the suction was fighting an air pocket. It wasn’t a cure for the curve, but it explained why readings looked off.

Login
Username:
Password:	Lost Password?
	Remember me