Vortex breaker 4

[1562]

What is the best design of vortex breaker?

 

[pmureiko]

I've seen designs with what looks like 2 layers of grating stacked on top of one-another and oriented 90 degrees apart above the suction nozzle.

In one case a distallation tower had a sawhorse left in it after turnaround. After the next turnaround the sawhorse was found and removed. Production dropped because it had functioned as a vortex breaker. This was listed in a Kister distillation book.

 

[Nosey]

In several 3 phase separators I have seen simple a capped + placed across the top of the liquid outlets. The height of the + was calculated to provide the equivalent hydraulic flow area to the outlet nozzle (a diagram would explain this so much better).

On one particular occasion we were refitting the vessel with some new internals and commissioned some CFD analysis. One of the points that raised was that for the new flowrates a simple 4 section (+) vortex breaker would not suffice as vortices would form in the quadrants. A capped 12 section vortex breaker was installed during the retrofit and working well when I left the company.

I can't remember the velocities or flowrates involved for that particular VB.

Hope that this is of use to you.

Regards

Nosey.

 

[pablo02]

On large vertical pumps (such a Circ water pumps for a power plant) they are basically two large plates at 90° welded into the suction bell.. similar the "+" mentioned by "nose" -- the center of the "+" at the center of the suction...

 

[scalleke]

Look at a heat exchanger type of design.

Take a larger diameter pipe and determine the correct thinner walled tubes that can be installet in the large tiameter pipe. Install the smaller tubes as a bundle in the larger pipe.

Attach the tubes to each other by welding.

We solved many a problem like this.

For better effect install longer section.

Best regards.

Scalleke

 

[Nosey]

scalleke,

If I understand you correctly the configuration is similar to:
---
/000 \000/
---
(apologies for the poor drawing).

Were the gaps between the tubes filled, or were they left as gaps to allow fluid flow?
Do you size the number of tubes etc on hydraulic diameter i.e. an equivalent diameter to the outlet nozzle.

When calculating the pressure drop due to the above configuration do you work out the entrance and exit losses for each tube section installed?

 

[scalleke]

Gaps are not filled.

Liquid passage should be greater than outlet nozzle.

I would not calcullate pressure loss but make one and test.

 

[BRT549]

The crossed plate design mentioned by Nosey can be refined by a rule of thumb : the length of the plates is 1.5 times the diameter of the outlet piping (kudos to TD2K). For instance, if you have a 12 inch outlet pipe, make the crossed plate 18 inches long. I have had good luck extending part of the plate into the outlet pipe itself by about 1/3. Using the previous example, the cross would be extended about 6" into the suction pipe.

reference a previous thread 794-29712 from last August for further discussion.

There is another design that is almost as simple, a baffle plate. Place a round baffle plate a distance of d/2 above the outlet pipe of diameter (d). The diameter of the baffle plate would be 4d. This simulates a larger intake area, thereby reducing the vortex and air entrainment.

 

[iceguy]

I have a centifugal pump side mounted to a shallow reservoir pumping cold water, 8" suction, 2000 gpm. I am anticipating a vortex so I extended the suction pipe through the reservoir wall by 4" so that I could attach some type of vortex breaker. Any suggestions on how I might achieve this with this particular configuration?

 

[BRT549]

make a cross 12" long by 8" wide. Basically take (2) 12" x 8" pieces of metal, make a notch halfway through each one about the same width as the sheet metal is thick (1/4" plate would be fine), and nest them together to make the cross. Take the cross and put it about 4"-6" into the suction pipe and tack it in place.

The cross should be the same type metal as the pipe. I have even made these with HDPE.

2000 gpm through an 8" pipe is over 12 feet per second. Any way you could go to 10" suction to stay under 10 fps?

 

[1562]

Thanks a lot from ur replys to my question. I will look to it and try it practically in future.

regard
1562

 

[DieselGuy]

ANSI Pump Standards, 1998 Edition Standard for Pump Intake Design, indicates that the use of a (1.5" bars) grating type cube, 3 inlet diameters on each side and 6 inches minimum above the inlet pipe provides a good vortex suppression for Centrifugal Pumps. They also provide several examples of guide vanes located at the inlet to break-up any rotational energy at the inlet.

I found this standard when trying to determine significance of vortex control for use in an eductor suction. I have not used either design myself. Hope this reference helps.

 

[PUMPDESIGNER]

I did not have time to study the solutions offered so I was staying out of this. However I think that one of our solutions would perhaps be useful.

We put little vertical walls from the floor upwards to the surface, about three of them surrounding the inlet nozzle. This prevents the body of water from swirling. These walls do not have to be tight against the inlet, just in the surrounding area to stop the water from moving horizontally.

PUMPDESIGNER

 

[smckennz]

Anyone sucking 2000GPM through an 8" pipe from a shallow reservoir is asking for trouble. I have tried the crossed plates on suctions and other flow straighteners, and I think they work about as well as a sawhorse; sometimes they do sometimes they dont.

I use S = 0.72*V*(d)^0.5

The 0.72 drops to 0.54 for smooth symmetrical entries.

S is the level of the top of the pipe below water level (or submergence)in metres

V is the pipeline velocity in m/s

d is the pipe diameter in metres.

Havent had a failure - yet.

Cheers

Steve McKenzie