8 replies to this topic

[Olaniyi]

Hi all,
First time on the forum! Hoping for many many future heathly discussions and arguments!

Question: Is it possible to specify a pump for a hydrocarbon liquid close to its bubble point? What kind of pump would you specify?
My gas processing plant has a skimming vessel into which foam + hydrocarbon is drained from an absorber column. I'm trying to provide a pump to take the contents of this skim vessel to a 3-phase separator some 100m. The problem is the skimmed "liquids" are typically at about 50-55oC and thus the hydrocarbons (mostly heavies) which get drained into it are at or close to bubble point. Due to the subsequent high vapour pressures, my NPSHA becomes negative.
Do I need a cooler, or would an inducer help? How about a Can pump? (I was originally thinking of a centrifugal pump but I don't think that will work)

Hoping for some insight! Thanks!

Ogeds (Newbie)

[Zauberberg]

Ogeds,

There is nothing unusual or highly specific in your application. There are hundreds of thousands pumps which are pumping fluids at their bubble point - 95% of process plant pumps are of that kind. You should worry only about selecting the proper type (recip or centrifugal), and providing adequate margin for NPSHa. For a fluid at its bubble point, the NPSHa comes only from the liquid head above the pump centerline.

For skimming vessels, I have seen usually reciprocating pumps (on-off operation) being selected for that particular service, due to uneven-intermittent (and relativley small) flow of skimmed hydrocarbons.

Suction-side cooling is a potential trap: you have to take care about cooler pressure drop vs. fluid vapor pressure reduction - that's te bottom line. If it is a fouling service (and it looks like it is), coupled with low fluid velocity inside the cooler (gravity flow), I wouldn't recommend you to go for that option.

[JoeWong]

Ogeds,

Welcome.

Something surprise me is you have negative NPSHa for a skimming vessel. Are you having pump center line higher than minimum liquid level, suction line too small, liquid too viscous ?
How did you calculate NPSHA ?

You can not have NEGATIVE NPSH A. Furnish us with basic information or submit your calculation for further evaluation.

Inducer will assist but won't solve your problem.
Can pump... may help but some operator and maintenance people may not like it.

For low NPSHA (but not negative) service, may consider progressive cavity pump, can pump, barrel pump, vertical submerged pump, etc.

[Olaniyi]

Thanks for the comments!
Let me furnish with basic info:
Max/min liquid level in vessel = 1.8m/0.6m
height of vessel above pump centre line (proposed)= 1.5m (vessel is sitting on a skirt)
Piping is 2" diameter, flowrate is estimated at about 3.4kg/s
operating temperature = 55degC
Vessel has an atmospheric vent, thus pressure = 1atm
hydrocarbons mixture (C3/C4/C5+, some amine, trace C02)
My main concern is the vapour pressure of my liquid mixture at that temperature

[djack77494]

ogeds,
I seems to me that you will have NPSHa of approximately 2.1 m (=1.5 + 0.6) less frictional losses in your suction piping. What you are describing and (erroneously) refering to as zero NPSHa is instead that you have a saturated liquid. The vapor pressure of the liquid = the imposed pressure. That's fine. If the suction flange of your pump was sitting right at the elevation of the top of your liquid, then you would have zero NPSHa. But your pump centerline is at least 2.1m below the liquid level. Thus you have at least 2.1m of NPSHa (less frictional losses, which look to be quite small). If your liquid is less than fully saturated, then you will have even more NPSHa! BTW, if you are storing a liquid hydrocarbon with vapor pressure = 1 bara or 14.7 psia at operating temperature, then you need a tank with a higher design pressure. Check into things; I think you'll find your liquid is not truly saturated and that you have a reasonable amount of NPSHa.

[JoeWong]

The worst condition is the suction fluid is at the boiling point in the drum where vapor pressure of the fluid is same as operating pressure of the vessel (ATM).

Your suction line is too small which possibly give you large pressure drop and lead to negative NPSHa. Consider to increase to 3" suction line. It will significantly improve your NPSHa.

[Olaniyi]

Exactly my thoughts Joe Wong. When I flashed pure C5 at 54barg and 55degC to an atmospheric vessel (in PRO-2) the resultant temperature was about 35-35.5degC. C5 boils at 36.5degC at 1atm so essentially I'm already close to my bubble point temperature.
I considered increasing to a 3" line but the other consideration is the exisitng line from which the tapping is to be done, which is a 2inch line. The vent which is being converted to the discharge line is also 2". Thus this would mean first expanding and then reducing the line. Also, would expanding to a 3" line just before the pump suction really result in any significant reduction in pressure losses along the line?

[JoeWong]

Can you pls sketch the suction configuration ?
Is there any drain connection available ?

[Olaniyi]

Thanks Guys,

Zauberberg, from your experience, can you give me an idea of the kind of specs for reciprocating pumps used for skimming vessels. In my case the amount of skimmed liquid is about 5-10litres and the discharge vessel is about 212 m away (with 18 90-degree elbows according to the project design specs being considered). The existing line to wich the pump is to be connected is 2 inches long.

With the typical amount of level obtained in this vessel representing about 18cm of height, I'm a bit nervous about putting the pump
Thanks a million!

Ogeds