Pump for circulating hot Limonene liquid 1

[Orionstar]

I need to find a pump that will circulate hot Limonene to create flow within a heated vessel. The Limonene will be at 80C and in a vessel that has been pumped down to 1 psia. It will be just under it's boiling point (VP at 80C is just over 1 psi. Properties of Limonene are; Sg .84, 0.9 Cp (at 25C, don't know at higher temp). Will be drawing from bottom of vessel and putting back in at 8" higher. Flow should be in the order of 5 gpm. Stainless Steel and Viton are good compatible materials. Any suggestions or education would be appreciated.

 

[EdStainless]

Can you put the pump below the vessel?
Maybe a air operated diaphragm pump, running very slowly with a large diameter intake.
If you pull any suction the solution will boil.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[MikeHalloran]

IF your vacuum pump has excess capacity,
AND is protected from limonene vapor by construction materials, cold trap, filters, etc.,
AND if the limonene will not be degraded by a modest flow of air through it,

you could try a sparge pump.

I.e., insert a stainless tube through a gland in the tank top, extending to nearly the bottom of the tank.
Leave the top of the tube open, or restrict it with a needle valve, etc. to control the flow of air into the tank.

The rising bubbles will circulate the liquid.

To make it more effective, surround the tube with an inverted funnel, small end up, which localizes the upward flow, and will more effectively stir the tank.

To make it still more effective, fit the bottom of the tube with a showerhead/ garden sprinkler/ perforated plate, pointing up, so that a cloud of tiny bubbles rises instead of an irregular burp of air from the bare tube end. (I use pneumatic mufflers for this, but they're commonly made of polyethylene, which may not be happy at 80C.)

No moving parts other than the air.
One new penetration.




Mike Halloran
Pembroke Pines, FL, USA

 

[Orionstar]

Thanks for the replies so far.
I can place the pump as far as about 2 ft below the vessel bottom. I guess that will get me a little static head for the pump inlet. I don't have air available for an air-diaphragm pump, but maybe slow motor driven diaphragm? How about an oversized vane pump run at a low speed?
Sparge is a good suggestion but pumping air thru the Limonene is not good for this as I believe it's oxidization creates unwanted compounds. Could use N2 but that is not readily available either.
maybe I should look into some kind of mechanical stirrer (shaft thru a seal in the tank cover?

 

[itsmoked]

A slow stainless propeller driven from above?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[moltenmetal]

Any pump you use will suffer from an NPSH problem because your fluid is boiling, or very nearly boiling. You only have 2 ft of elevation head plus whatever head is in the vessel. It is possible to find a small low speed centrifugal pump with an NPSHR of only 2', but you need some allowance for frictional losses in your suction piping including the sudden entrance from the vessel into the suction pipe. Unless you have a few feet of liquid operating level in your tank, you will likely either need to raise the tank or drop the pump suction.

A piece of somewhat counterintuitive advice about a centrifugal pump if you choose to go that route: choose one with a full-sized impeller. The largest impeller size for a particular volute gives the minimum NPSHR. So what you're looking for will be a sub-ANSI centrifugal pump suitable for that low flow (5gpm), operating at low speed- you don't need much differential head, so that won't punish you much. Pumps made by people like Magnatex, March etc. come to mind. Under vacuum, you'd want a magdrive in my opinion, not a single mech seal.

An air diaphragm pump won't be damaged by the boiling which occurs in the pump suction- it will suffer from a capacity loss though. Unfortunately you can't really modulate the rate at which an air diaphragm pump's suction stroke takes place- you can only really change the stroke frequency. They're not a good choice for systems already operating under vacuum anyway. A slow moving motor-driven diaphragm pump might be a better option.

A mixer won't suffer from those problems and is likely what you really want.

A sparger isn't a practical option in a vessel under vacuum containing a boiling fluid.

 

[Orionstar]

Thanks for the info mm.
A mixer or propeller is off the table at this time.
Also, there is no air available for an air diaphragm pump.
Good news is the process vacuum requirement is much reduced - now we will probably run the vessel at around 300 mmHg. Also, the flow requirement might be as low as 2 gpm. So I think I should be able to use a slow speed centrifugal pump. (BTW - my liquid operating level is only 10 inches, so my total head from the surface to the pump inlet is no more than 32 inches.)
I have thought mag drive would be good also, but my inquiry to Iwaki (who I have used in the past for other apps)led to them not recommending their mag drive pump (at least the small ones) in a sub-atm pressure due to the balance of pressure on the pump impeller being forced the wrong way (in a direction that did not have a bearing surface. Do you know if this is true for other manufactures of small mag drive? Or maybe with 2 gpm I should just fins a slow moving electric driven diaphragm.

 

[moltenmetal]

Never heard Iwaki's concern- bet it's very specific to their design. Most pumps have bushings on both ends of the impeller and magnet assembly so they can handle thrust in either direction. Bet plenty of Iwaki pumps are operated with substantially sub-atmospheric suction pressure- maybe they just don't last. Talk to March and Magnatex. March makes a line similar to Iwaki, whereas Magnatex (made in Japan- may have a different brand name outside North America) are really good quality industrial units but built at the sub-ANSI sizes you'd need.

A motor driven diaphragm such as a Hydracell is another option as long as you have compatible elastomer materials and operate it slow enough. They can only handle limited temperature but I think you'd be OK there.

 

[georgeverghese]

Another option may be to cool this limonene down (with a low dp HX) in the pump suction line to some temperature such that its vapor pressure is adequate to provide some 6ft of NPSH or so to enable operation of a recirc pump. See if the heat input to this vessel is adequate to bring the temp back to 80degC.

 

[moltenmetal]

Cooling the suction line is a good strategy in situations where the liquid circulation is not heat transfer as stated by the OP.

 

[jari001]

I would recommend a peristaltic pump, but you'd have to find the right tubing for Limonene.
(Edit: Removed all those awful blank lines, where'd they come from?!)

 

[moltenmetal]

Peristaltic is a bad suggestion due to the fact that it's operating under vacuum, even if you could find a suitable tubing material for handling hot limonene.

 

[Orionstar]

I spoke at more length with Iwaki. Turns out the negative pressure worry is due to the thin plastic shell that separates the impeller magnet (in fluid) from the outer motor magnet. With enough negative pressure in this area, the shell can deform inward and rub on rotating impeller magnet. We agreed that as long as we have some positive discharge pressure, for example by throttling the output line and not as much vacuum in the chamber we are pumping back into, the area behind the impeller more closely matches the discharge pressure and should be OK. Another thing I learned is that very few manufactures of small centrifugal pumps publish or even know what their NPSHr is. The best Iwaki could do was a statement of "6 - 10 ft of suction head". We may go ahead and try this pump as long as I can run the process chamber at 600 mgHg or so.

 

[moltenmetal]

Thanks for the feedback on the Iwaki- and yes, they have poor info on NPSHr. Magnatex definitely knows their NPSHR values accurately.

 

[jari001]

Not to hijack the thread, I just want to understand moltenmetal's rebuttal to my suggestion. Is your reason to say the peristaltic pump is not a good choice because the suction it generates may cause the Limonene to flash as the liquid tries to flow through the tubing?

If that is the reason, I will say that I have been able to use peristaltic pumps in vacuum conditions but I will admit that the inlet tubing submersion to maximize static head is is very critical and not at all forgiving.

 

[moltenmetal]

Peristaltic pumps rely on the "springiness" of the tubing to generate the required suction, given that their housings themselves (assuming the OP would be using a peristaltic with a housing for the hose) are at atmospheric pressure. The necessary force at the suction is generated by using very heavy wall tubing, which must be an elastomer to work. But given that the service is not only at nearly full vacuum but also HOT as far as those hose materials are concerned, and uses a medium which will swell the normal hose materials, it's a poor service for a persistaltic.

Can you use a peristaltic as a crude vacuum pump at low temperature with nonhazardous fluids? Sure- you can also use them to pump mixtures of gas and liquid etc.

 

[MikeHalloran]

If the tank shell can deal with a little localized heating,
AND the limonene doesn't decompose when boiling,
a percolator pump could work ok.

Comprising, e.g.:
- a heater raising the temperature of a small area on the bottom
- a shallow cup, covering the heated area of the tank shell
- a standpipe joined to the cup, and extending upward a bit,
- cup and standpipe loosely constrained so they can rise together, but not translate laterally
... in operation, the limonene within the cup will star to boil,
a vapor bubble will appear and lift the cup and standpipe,
then 'burp' out of the standpipe,
and the cup will refill with less hot limonene as it falls.



Mike Halloran
Pembroke Pines, FL, USA

 

[EdStainless]

In reply to Iwaki comments, find a competitor that uses a metal shell.
Do you have pure nitrogen available? Use an eductor and N to make a jet pump to stir with.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Orionstar]

To close this out, as the OP, we never did build this job (for other reasons). We had settled on using a March mag drive pvdf pump with the vessel at -2.5psig. With a discharge throttle valve we expected to have a positive discharge pressure (we only needed 1/3 of the wide open flow rate) so that the pump chamber magnet separating wall wouldn't want to "cave in". With a wide open and flooded intake we should see enough NPSHa for this pump. Again, never got to try it though. All the stainless mag drive pumps were too expensive. No N2 available and didn't want to introduce anything to the vessel. MikeHalloran idea interesting, but not going to give the circulating flow in the vessel that they wanted. I was going to suggest a simple stirrer with a good shaft seal but we never got that far. Thanks for all the input.

 

[itsmoked]

Thanks for that. Happens a lot.

Keith Cress
kcress -

http://www.flaminsystems.com