Cavitation in centrifuge pumps

It is highly advisable to have somebody inspect the plans and do proper calculations.

Alternatively you should give more detail.

Temperature of water? Flow rate? Suction pipe diameter? NPSH of pump? etc etc.

Assuming that the level of the water in the cooling tower sump is kept above the pump suction level then this is not an unusual design. As to whether you will get cavitation or not, it is a matter of supplying the information as suggested by Hendrik.

Probably, and quite probably at some point you will flood the cooling tower. As Hen the cat suggests, have a local pump expert carefully review the plans and proposed setup and design the problems out before construction starts.

Cavitation is certain if the system is gravity fed.

First you will flood the tower, then you will have cavitation problems for a while...

I do not entirely agree with your sequence, it really depends on the system set-up and operation.

I have seen cooloing towers start up and operate successfully with the pump suction elevated 1 to 1.5M above the cold water basin level. The trick is in the details.

You are right. If losses through the suction pipe are low, temperature of the water isn't too high, and the NPSHR of the pump is low - there should be no problem with a 1 to 1.5m lift.

We design & install cooling systems every day from 10kW to 100 MW 2"NB to 48" pipe

As others have said - it depends in part on the liquid level in the sump/base of the cooling tower, and additionally the line losses to each pump - not all suction lines are created equal, the water temperature - warmer water is more prone to cavitation, and the design of the pump - not all pumps are created equal.

If you employ a local pump person, you need to be certain he has only your interests in mind - i.e. don't count on the guy selling you the pump to give you the whole and accurate picture. Beyond the sale of the pump - he's also looking forward to selling you repair parts and services.

Good Luck.

You can refer this to the pump manufacturer who can confirm whether your pipe size layout etc is in order and there will be no cavitation. Many pump manufacturers have computer simulation and can easily check this.

Don't create a design problem, put the pumps imediatly adjacent to the cooling tower in a flooded suction manner.

You are requested to provide the following data at least :

a. Piping isometric for suction line with all dimensions and sizes, showing all fittings, flanges, elbows, valves, .... etc.

b. Piping material.

c. NPSH of the pump.

d. Water Flow rate.

e. All pump data.

Suction pipe line diameter is 350mm. Pump capacity is 375m3/hr. 3 pumps connected parallel, 2 will be in use 1 stand by. Water temp will be 35-38 deg c.

Good information. Additionally, what is the difference in height between the liquid in the bottom of the cooling tower and the centline of the suction of the pumps? What is the discharge pressure generated by the pumps, What speed do the pumps run (RPM)? Do you know the diameter of the impellers of the pumps? Am I correct in thinking these are single stage pumps? Do you have access to the pump curves?

The centrifuge pumps Ive seen are fairly solid. Only time cavitation would be a problem is if the pump was to far from the source (vertical distance, which also depends on viscosity) or the input line was to small. To predict this ask pump manufacturer for max vertical on the input. Keep the input line the same ID as the opening on the pump itself. I like to keep a slight restriction in the output on a flat run to keep the pump at a higher working pressure but if your going vertical you don't need that at all ... LOL