NPSHa with dissolved N2

[DJK2]

I was hoping someone on this forum may help with a question I have.

I have a condenser that runs off into a flash vessel. The vapour then goes off to a vacuum pump and the liquid to a gear pump. The flash vessel runs at 0.4bar and the vapour pressure of the solvent is 0.04bar.

The question is, how to calculate the NPSHa for the gear pump.

Should I just take the Vap Pressure of the solvent (0.04bar) or is the pressure in the vessel equal to the Vap Press.

Thanks for your help.

 

[ione]

If you have to deal with a flashing liquid the NPSHa is the height difference between the level in the flashing tank and the pump centerline minus suction circuit losses.

 

[JJPellin]

I am not aware of any method to accurately account for dissolved gases of the sort that you are describing. It definitely does have an affect in some applications. We know from experience that if we use nitrogen pressure on the barrier fluid tank for an API Plan 53 seal system, above a certain pressure, we see damage from flashing at the mechanical seal faces even if the vapor pressure of the barrier fluid is well below the operating pressure.

If you had unlimited time and money, you could perform tests to determine the effective partial pressure of the dissolved nitrogen. If you had the luxury of extreme conservatism, you could assume that the liquid was at vapor pressure at the surface and only account for elevation head. In reality, the truth lies somewhere in between the vapor pressure of the liquid before it contacts nitrogen and the elevation head with no allowance for vapor pressure margin.

The pressure you describe seems low to me for this affect. We start to be concerned at about 150 psi in a seal pot. For a nitrogen blanketed vessel at less than 1 barg, I think that the affect on NPSH would be negligible. If you want to be slightly conservative, increase you required minimum NPSH margin by 1 or 2 feet.


Johnny Pellin

 

[DJK2]

Thank you for your replies.

Although it will mean a change to the layout we still have the option of increasing the static height to the pump so I will propose that we only account for elevation head to be on the safe side.