Cryogenic Pump Priming Problem

Sounds like heat is finding its way into the liquid from an unexpected source in sufficient quantity to vaporize some of the liquid. Maybe internal rubbing contact between impeller and casing or excess heat generated in the pump bearings transferred into the pump shaft and then to the eye of the impeller.

Ed Weldon

He's probably right, but don't forget pressure vs boiling point. If flow rate gets to high, can your input pressure be dropping causing vaporization? Just a thought, maybe not a good one.

Physically, either one of two things or both. As Ed previously noted, heat input could be coming from somewhere, and cause some liquid O2 to convert to gas. The other possibility is that the static pressure is somewhere dropping low enough to convert to gas.

At -185 celsius, the saturation pressure for O2 is 0.0814 MPa. Is the operation close to that? Might also check seals to see if there are any signs of excessive wear would indicate unintentional heat generation that might raise the fluid temperature near the seals.

For more info on the cryogenic properties of O2 (or other fluids), check out

http://webbook.nist.gov/chemistry/

Under "models and tools", choose "thermophysical properties of fluid systems", then choose the fluid, desired units, and saturation properties to get a table of pertinent fluid properties. To prevent cavitation, static pressures in the fluid have to be kept higher than the saturation pressure at a given temperature, or the temperature kept below the saturation temperature at a given pressure.

If you can, thermocouples and pressure taps at the inlet and exit of the pump might yield some insight into what is happening, and lead to a solution to better control the fluid conditions.

Trend the problem: Has it always been like this, or just recently? What has changed? Process conditions changed? Operators? Pump recently serviced? Physical changes to the pump (inducer removed, etc), tank changed, NPSHa changed? I have some LNG pump background and we cool LNG pumps for 3-4 hours, not 30 minutes.

Agree with you PetroPower, 30 minutes is very short unless this pump is very small. Cryogenic pumps and the system in which they run need to be at operating temperature before starting. This usually means vents from the pump back to the vapour space of the vessel and bypasses around the NRV's in order to create backflow and cooling through the pump. Check the level of the suction vessel and also make sure that the vortex breaker is in place. Get ops to increase the minimum level of the suction vessel if there is not adequate head. If you have a suction strainer check to see that the startup mesh has been removed. It is not required - not ever.

How old is this pump?.

What type of pump is it? Centrifugal? Piston?

Has the problem only recently started?

If suction conditions have changed, then some bubbles could form in the suction, leading to cavitation. A cavitating LOX pump can be dangerous!

If discharge conditions have changed and you have a centrifugal pump, low flow can cause recirculation within the impeller which can generate cavitation.