Cryogenic piping design

[Erick21]

Dear All,

I'm looking for some information on how to design a suction piping for two centrifugal pumps connected to a common header and a suction tank. When both pumps run together, both run fine. As soon as you stop one pump, the second pump starts to have some problems after sone minutes and completely loose the suction. It seems the suction line of the other pump when it is stopped has an influence to the other one. Is there somebody who has already seen this phenomena. Liquid is ethylen at -100°C

thanks

Eric

 

[iainuts]

Hi Eric. Most likely the piping on the inlet of the pump that isn't running has liquid in it that's boiling off. The resulting gas is pushing liquid back toward the T where the suction line splits and into the running pump. Eventually, the gas begins flowing into that T and into the running pump which results in cavitation and loss of prime.

There are various things you can do, but before reviewing those, a couple questions:
[ol 1]
[li]Do you have any insulation on the suction piping?[/li]
[li]What diameter is the pipe?[/li]
[li]What flow rate are the pumps?[/li]
[li]Can you attach a photo of the suction line?[/li]
[li]A sketch of the inlet piping showing valves, a vapor return (if you have one) and any liquid traps would help.[/li]
[/ol]

 

[Erick21]

Thanks for your feedback,

For sure there's insulation on all suction piping. It is almost 2inches thick in piping of diameter 3 inches.
Pump flow rate is about 10 m3/hr.

I will try to have more details of the piping isometrics, insulation caracteristics,..

I was just wondering if it is a know problem for cryogenic piping design to NOT pipes pump to a common header. I'm sure that the insulation calculation for each pump was calculated using only the suction line of the pump itself not the other lines.

Eric

 

[1gibson]

Not a complete description of the system, but here are a few guesses:

1) Unless both pumps have check valves on the discharge, just "turning one pump off" will cause problems. You would have backflow through the idle pump, causing high flowrate of the running pump, recirculating some of the fluid through the working pump will heat up the fluid and possible exceeed vapor pressure. The high flow will mean higher NPSHR, possible cavitation.

2) Delete the backflow and the recirculating part of #1, and the remainder could still be problems. High flowrate due to not enough system resistance, higher NPSHR, possible cavitation.


Need a system sketch, pressure info (including discharge and suction pressure readings, and value for vapor pressure @ relevant fluid temp) and control logic.

 

[iainuts]

Piping 2 pumps off a single header is a common practice, but you'll need a way to isolate them.

The amount of insulation is relative. I've seen pumps your size with no insulation on lines 10 to 20 feet long.