Is there any problem in using a rotary lobe pump to transfer butadiene (lpg)? 3

[lcms]

Dear Sirs,

I am working in a project where we will have to transfer a liquified gas (Butadiene, 3 kgf/cm2g; 30C) from a vessel bellow the sucction pump to charge a reactor that requires at least 7 kgf/cm2g at pump discharge.

I am considering to use a rotary lob pump I have available at my WWTP (MBR permeate) that work well in a self prime mode, but I am concerned since I have never heard about this kind of pump in LPG transfer.

Could you please explain me why?

Thanks.

 

[Latexman]

I prefer a sliding vane pump in buta service. Rotary gear pumps wear and lose their tight clearances continuously. Their performance is continuously degrading. Spring loaded vanes allow sliding vane pumps to keep their performance more consistently. Both pumps need PM, but sliding vane pump performance is much more constant.

Good Luck,
Latexman

 

[lcms]

Dear Latexman,

First, thanks for your advice. It is exactly what I would do if I had do specify a new pump, but I am trying to take one of ours available pumps to make the project feasible.

Since I have to pull the butadiene from a vessel underground, I have to consider a sel priming pump, what the sliding vane really is (the best).

That is why I have considered a rotary lobe pump we have as a spare in our WWTP. However, you are completely right considering the rapidly wearing in this application (low viscosity, two phase flow) and I am also worried about this option.

Now, I am think about two options:

1) Install one of several can centrifugal multistage vertical pumps (from projects suspended in past) stright in the vessel;

2) Pressurize the vessel to prime our existing horizontal centrifugal pump that work very well in transfer Buta while there is liquid in it.

I would appretiate if you could give me your opinion about that.

Best Regards,

LCMS

 

[Latexman]

A wise old Project Manager once told me, "It usually costs a little extra, to use used equipment".

Good Luck,
Latexman

 

[georgeverghese]

Be careful with the electrical area classification limitations with that WWTP pump - speak to your plant electrical engineer. Same goes for the shaft seal / flush plan - believe 1,3 butadiene is a known carcinogen.

 

[TiCl4]

To clarify, do you have a gear pump or a lobe pump? A lobe pump will not have the issues described by Latexman, as it relies upon clearances, not gear enmeshment, for providing positive pressure. Lobe pumps do not suffer from wear like gear pumps do if the fluid is clean. To put a further point on georgeverghese's point, you will need ensure your pump motor is rated for a Class I Div I area (maybe Div II, depending on your area electrical classification).

You will also likely need to provide a relief line on the pump discharge in the case of pump deadhead. I'm not familiar with how "dirty" butadiene gets (do the residuals on the pipe tend to polymerize and form grit/skins/etc?). If it is a fouling process, you'll also likely need to consider a rupture disc instead of a PSV for relief.

Lastly, sealing may also be an issue. I have used double mechanical seals with a water flush on monomer transfer pumps (with water pressure higher than line pressure), but the application was emulsion polymerization where a little bit of water contamination was not an issue. Again, georgeverghese is right - you will need to select your sealing arrangement carefully to avoid the possibility of leaking.

Also, the conditions you provide (~300K, 3 kgf/cm2) seem awfully close to the boiling point of butadiene (based on extrapolating Antoine constants from NIST). How much prime is needed? Trying to self-prime a liquid that is near its boiling point is difficult, as the liquid may flash when being pulled. I don't know the dynamics (evaporation rate vs volume draw from the pump), but it may be impossible to self-prime without pressurizing the tank more. You may also wreck your seals when trying to prime if it runs dry too long.

Lots of challenges exist with this retrofit idea. Option (1) provided in your second post seems like it would be more feasible and less expensive than trying to retrofit a pump that isn't purpose-built for the application.

 

[lcms]

Dear TiCl4,

First thank you for your useful post.

One of my available pumps is a Borger LOBE pump (not Gear) model PL-300 and probable would not have the issues described by Latexman concering weraing, but as you have noticed in your e-mail, the butadiene is very close to saturation condition and if I would use this pump it could bring me serious cavitation problems that I was thinking to solve pressurizing my vessel with Butadiene gas but, even so, it could have been some problems in the future.

Therefore we have decided to go further with the vertical centrifugal multi stage CAN pump (Whorthington model 6OEA/R/4C) and forget the Lobe pump. You are compltely right concerning our pump motor that must (and will) be rated for a NEC Class I Div I (GR.D) area.

The destination vessel of the Butadiene has already a PSV with a rupture disk (relief line on the pump discharge) but thanks for warning mefor risks related to discharge pump dead end.

Best Regard,

Luís.