I've seen deep well centrifugals used for similar applications for handling the condensate from a surface condenser so it's not obvious to me you'd immediately pick a PD.  That said, PD is also another choice, like most applications, it's a matter of what flow and pressure you need.  

What is the fluid?  What flow rate do you need?  Is the flow rate fairly steady, do you need control, continuous operation, intermittent?  

Its pretty low flow and pressure. Usually no more than 30 gpm and no more than the 10 psig. The fluid is recovered groundwater and the flow is intermittent (controlled by level switches in the tank), only minimal control is necessary (manual control valve or bypass vavle).

For this small a flow, you may want to consider a regenerative turbine type of pump.  (You may want to check www.rothpump.com to see an example, and incidentally, I do NOT have any relationship with Roth Pump.)

These pumps can be quite forgiving on NPSHr, but you would do well to make sure that the pumpage does not include significant abrasive material.

If abrasives are present, then you may be better served by a progressing cavity pump (Moyno or seepex are examples, and again, I have no relationship with these companies.)  These are positive displacement pumps, and your system design should take this properly into account.

Thank you for your help. I have had good experience with progressive cavity, but have not used regenerative turbine - this I will look into. But what about my original question - can you use an air-operated double diaphragm pump to pump out of a vessel under vacuum? Your comment on this would be appreciated.

I'm not aware of any air operated double diaphram pumps that are intended for drawing from a strong vacuum.  Most are intended for suction lifts of about 10 ft water column.  Even if the pump can draw from such a strong vacuum, I would expect short diaphram life due to the severe flexing that these conditions would impose on the pump's diaphrams.

Another potentially important issue would be the cost of operation of a compressed air powered pump.  Compressed air is normally a very costly energy form.  The energy required to compress the air is generally much greater than the energy required to directly drive a pump.

Please be very careful when selecting a progressive cavity pump for abrasive fluids.I've had very bad experiences with this type of pump in fluids that contained relatively small amounts of abrasive particles. Stators might last only as long as several weeks, which makes them expensive to operate and maintain. Has anyone else found this to be the case?

Hi RV.

As far as I know the diaphragm pump could "suck" up to 21'
or 6 meter.
Your tank is far below it , about  30' or 9 meter  .
Even to a good one centrifugal or regenerative pump it will be a hard task to do .
There is any way you could break the vacuum , and for a while  use a conventional pump ?, as you told that the  flow is intermitent it could be possible..
I also thoug on a eyector , but it need at least 25'.
So please try to change some thing.
Further more : why the tank is at such vacuum.?

Other question , what do you mean with: Knockout tank???




  

Pardal

By knockout tank I mean a liquid knock-out or vapor/liquid separator tank. The tank is under vacuum because it is upstream of, and in line with, a high vacuum pump (liquid ring) used for dual-phase extraction of soil vapors and groundwater from wells. The soil vapors and groundwater need to be separated for separate processing, as well as to protect the vacuum pump.

RV .

More questions:

Is the dual phase a continous process?
Could it be stoped, for a while at least?
Could you "rise" the abolute pressure , for a while, just to allow a pump to "suck" it?
Or maybe you can use a old reliable piston pump.
But remember that only have a few absolute pressure.   
You shall consider that if the tank is at absolute 0 , you can not take out any amount of any thing from it.
Because ther is no pressure to force it out.
Hope it help (HIH)

Pardal

How do I relate the published data about double diaphragm pumps and their "dry lift suction capacity" to NPSHr? The manufacturer catalogues I have don't show NPSHr on their diaphragm pump performance curves. I presume I have to convert the dry lift capacity they publish (10 ft dry suction lift capacity, for example) into a number that I can compare with the NPSH available to see if the pump is suitable.

Wouldn't I experience problems with the piston pump too if there are some abrasives/ small solids present due to the internal valving?

I guess it looks like a progressing cavity is still the best way to go. This is what we have used in the past but I find we replace the boots (stators) quite often and I was hoping for an alternative, espescially where compressed air is available (I realize the costs associated with the compressed air now, however).

Thanks for all the help.

RV
Normally all NPSHr , stand for Net Positive Suction Head requiered , are for water , clean , at 20 C degree.
Dry , means no water inside the suction pipe.

Your NPSH aviable shall be less , form atmospheric press, than the NHSPr.

---------------   atm press
 

---------------   NPHS avaible Your tank press


---------------   NPHS required  your pump capacyti normally no more than 7 meter H2O
 



---------------   0 absolute press -or 10.033 mt H2O or 30" HG

Pardal

RV
I do not know nothing about progressive pumps , only the name, I never work with it.

Pardal

RV,

I have only had a quick look at the posts above, so I apologize for any duplication of answers.

First I would like to suggest that there is no relationship between a pumps capability to prime and NPSHr.  The ability to prime is dependent on the pumps ability to move air out of the pump discharge (thereby creating a vacuum which is filled by air (gas) from the suction side of the pump).  NPSH is related to liquid flow and the affect of liquid pressure drop in the pump suction due to the high velocity of the liquid in the pump impeller eye section.

Another potential solution to your pumping requirement, is to install the pump below the knockout tank - thereby providing sufficient head for NPSHr.  If the knockout tank is located at ground level, can you install a simple Vertical Turbine Pump (VTP) in a long suction can.  With a 10' elevation from the pump to the bottom of the knockout tank you probably would eliminate the NPSH problem.  Use large diameter suction piping to minimize friction losses in the suction pipe.

This is a typical installation of condensate pumps in power plants.  If this type of solution is possible, reliability of a VTP will far exceed PD pumps.

Hope I havn't repeated this suggestion if recommended by others.

Have you thought about using a hose pump or similar peristaltic systems?  If you can live with pulsating flow, these pumps work extremely well for applications with low NPSH.

RV,
Double diaphram (Wilden) pumps are not without their own problems.  They're noisy, pretty ineffecient, and prone to diaphram failure.  I use them when there is NO other option (e.g., remote locations without electricity that require auto-start).  Even then, the alternative to Wilden is often blow cases.  These are separate vessels that sit below the knockout drum and have a check-valve in the drain line from the drum to the blowcase.  When the blowcase is full, control valves shut off the blowcase-vent to the drum, open a blowcase-drain to wherever (tank, pipeline, etc), and open a power-gas line (e.g., you could use the air you were going to drive the Wilden with) to pressure the blowcase and discharge the liquid.  At the end of the cycle the power gas and dump close, the vent opens and the check valve allows liquid to begin flowing back into the blowcase.  I've used these devices to move 30 gpm from 10" Hg vacuum to 150 psig--you need a pretty big blowcase and quite a bit of power gas, but less power gas than a Wilden to move this volume.

From the discussion above, it looks like you've rejected centrifugul pumps so I'm assuming that you don't have any way to provide a hot well.  There are millions of steam applications with condensers on a very hard vacuum and a centrifugul pulling on a hot well.  No other technology has proven nearly as effective or robust in this sort of application.

With less than 20 psi head and 30 gpm, I wouldn't use PCP's or ESP's because of wear problems in intermitant high-demand service.  Also, I avoid PD pumps unless there is no alternative (I generally only use them for really high discharge pressures).

David

How about adding a sump tank below the knockout tank. When the sump is full, isolate it from the knockout, open it to atmosphere and pump.

RV,
Check out Krogh Pump Co. in Benecia, CA.  They make a flange mounted centrifugal pump designed for pumping from vacuum receiver tanks.  The pumps require very low NPSHa.
Good luck,
Piffer