Troubleshooting a 2 stage liquid ring vacuum pump

[jari001]

Hi everyone,

My situation is that I have 3 vessels that are all connected to one 2 stage liquid ring vacuum pump (Kinney, further info below) and this pump is used to pull methylene chloride (MeCl) from the batch. We are having performance issues such that the vacuum set points are not achieved and that it takes a much longer (2x more time) time to pull the MeCl from the batch even when we are lets say within 1 to 2 in. Hg on the set point. I am going to test the vacuum pump in idealized situations were I blind all other lines on the inlet header and try to pull down to the lowest set point on a vessel. I expect to have to hold a low pressure state for a moderate period of time to look for leaks if I cannot attain a set point. Do I need to provide something in the vapor space of the vessel for the pump to move to protect the pump or is the seal fluid supposed to* provide the needed cooling as well as make the pockets that the vapor gets carried in when inside the pump housing?

I am assuming I should provide something for the pump to actually move, but if it is not necessary then I can proceed with my test if ops folks are unable to charge some water to one of the affected vessels tomorrow; we are cleaning the water system tomorrow so I have a window to run a couple tests while the batches sit but I'd need ops to fill water before they shut that system down.

Thanks for your time,
J

Info gathered:
Lowest pressure set point on a vessel is 12" Hg
P&ID specs motor at 20 HP at 1000 rpm
Batch temp maintained at 26^oC
Seal fluid is methanol at 10^oC, I think it is constantly circulated from the separator part of the vac system to a HX for temp and then back into the pump
Pump curve and other info
Manufacturer's Manual
*I still have to confirm there have been no deviations on the seal fluid volume or temp.

 

[Compositepro]

Are you talking gauge pressures or absolute pressures? 20" w.c. is not much of a vacuum and you don't need a two stage pump for that. Perhaps you mean in.Hg? 20 "w.c. absolute pressure puts you under the vapor pressure of methanol at 10C, so the pump cannot get there.

 

[jari001]

I edited my post, the lowest set point is 12"Hg not 20" w.c. and other mentions of w.c. were edited to be Hg. Sorry about that confusion, they should have been in. Hg (vac) from the start. I am working in gauge pressure.

 

[EdStainless]

You do know fluid temps correct? Are they cold enough?
That is issue #1 with liquid ring pumps, you can't pump below the boiling point of the seal liquid.
If they are cool enough and give good vacuum blanked off, then hunt for leaks.


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

 

[georgeverghese]

Methanol sat vp at 10(??)degC = 55mmHg = 2.2inHg - so seal fluid is okay for 12inHg setpoint
Methylene chloride sat vp at 10degC = 4psi = 8.25inHg - so methylene chloride would be still be in liquid phase at 12inHg - this may be the reason why the suction pressure is hovering around 13-14inHg.

The vac pump final suction pressure is now a function of the operating temp in these vessels, and not following the setpoint.

 

[jari001]

@georgeverghese - I added the batch temp information, that was missing from my OP. At 26^oC, the vapor pressure of methylene chloride is ~500 mmHg (~20 inHg), so that checks out as an appropriate set point.

I believe I have leaks which is why I want to run these tests, my concern is coming from an equipment operation standpoint. I went ahead and asked for water to be charged to the average volumes each tank would hold in preparation for today's tests just to be safe. However, I would still like to know if it is acceptable to run this type of vacuum pump without having any vapor to remove from the vessel. I initially said I thought vapor traffic was needed (equating this vacuum pump to a centrifugal pump), but now I realize that compared to the seal fluid, any vapor that this pump is removing would do almost nothing for heat transfer so it would be okay to run the pump without a continuous stream of vapor. Please do correct me if I am mistaken.

Thanks,
J

 

[georgeverghese]

Okay, with 12inHg setpoint, and a sat vp of 20inHg at 26degC, that would draw out all the methylene chloride from these vessels.

Dont think there is any risk of zero vapor feed to this pump, since even water has a non negligible vapor pressure at 26degC, and more so for MeOH. Agreed, most of the heat developed in the pump would be absorbed by the seal fluid. Yes, can see the rational for these leak tests now, if you find the vac pump is struggling to get down to 12inHg even after all the methylene chloride is drawn out - or maybe there are some liquid phase non azeotropic contaminants in this methylene chloride with a sat vapor pressure lower that 12inHg ?

 

[jari001]

The process chemists and my colleagues have not indicated any known issue with azeotropes forming. I am waiting on info to make sure our pure MeCl vendor hasn't changed entirely or we're being a different grade (fishing trip essentially). First vessel tested yesterday (Vessel A, 200 US gal capacity, filled with 100 US gal of water) and the only observation I made was that whenever we changed the valve setup (i.e. opening and closing different valves on the vacuum header) the vessel would lose 1" Hg vac but then stay steady. I think I will have to test another vessel and see if the 1" Hg vac loss is truly indicative of anything.

 

[jari001]

I'm progressing with pressure tests, no leaks yet, and trying to setup time to have the pump rebuilt. I will likely have some more questions based on results (that will fit into other forums) but I don't have any further updates on the operation of this vacuum pump.