Reducers in Centrifugal Pump Suction

If you have 30-40 m more NPSH available than is required I don't see a problem, as long as you have a smooth transition from 24" to 28" and at least 10' of 28" pipe before the inlet of the pump.

I would be more comfortable having more complete information (pipe layout, dimensions, elevations, pump curve, etc.), but I'm with Lyn. And then, even if we have all estimated wrong, you might be able to restrict the pump discharge to eliminate cavitation if it occurs.

Crane Engineering has good information on fluid flow.

Crane is the goto standard when it comes to fluid flow information.

Lyn and Tornado,

Piping up to pump suction nozzle is of 24", as the existing pump nozzle was 24". with this new 28" suction nozzle pump, we intend to use existing piping just with a new reducer (24X28) in pump suction next to pump suction nozzle. If technically acceptable, this piping arrangement is most suited to site, as not much piping modification.

However, if 3D - 5D straight length requirement need to be fulfilled, we need to provide 28" pipe in pump suction, then connect it with existing 24" pipe with reducer.

so my question is again, if we have more than sufficient margin (6-7 times) on NPSH, do we still need to care about piping arrangement?

NPSH(A) or (R)?

https://en.wikipedia.org/wiki/NPSH

http://www.engineeringtoolbox.com/npsh-net-positive-suction-head-d_634.html

http://www.pumpfundamentals.com/centrifugal-pump-tips.htm

My feeling is that there are too many unanswered variables to give any answer other than, try it and see.

Some unanswered variables are site elevation, fluid termerature, density, viscosity and working pressure.

Try it and see.

Tornado will have his own opinion.

They've told you once, what more do you need to know?

You need to check the suction hydraulics to see whether it's OK. As said, if you have that much NPSH margin (at the increased flow) there should be no problem.

The new pump has begger suction port. Apparently, the new pump has higher discharge than the old one. If this is the case, 24" suction will prove to be inadequate. Check that the suction velocity in suction pipe is less than one met/sec. The suction line should be bigger than pump suction portto achieve this condtion. Even if you have positive suction head, only expanding by 24x28 at suction will not help.

If all the piping is 24 inch, the outlet is no larger than the inlet. This is not established, but inferred by me.

So the fluid velocity will likely increase which only adds to the likelyhood of turbulance and cavitation.

As already mentioned, too many unknowns here.

OP should be prepared to provide ALL pertinent information to pump man.

There is no need to repeat stuff in the original post.

How did the phone call to the pump supplier go? No joy there?

no brainer. IF the original is 24", and output is 24" (even with the new and improved pump), going to 28" suction should help not hinder.

I cannot tell you from where I sit if this is appropriate for the pump curve being employed, as you failed to provide all the needed information.

"you failed to provide all the needed information."

Should be: you failed to provide all the any needed information.

True. So did I.

You want someone to give you permission to take the path of least resistance.

OK, go ahead.

I'll give you a full refund of your retainer is if fails.

If you're just gonna drain one swamp one time, maybe none of this matters....

Sounds more like a flood situation....that's a big pump...

NPSH_A versus NPSH_R may be paramount, but there are further considerations.

Even if the suction piping can deliver overall adequate flow/pressure to the pump, turbulence may create local zones of less pressure, and hence cavitation. This is what underlies the commonly seen criterion of say 6x pipe diameter of straight run leading to the pump.

Fluid pre-rotation is sometimes another issue. At least one manufacturer, Cornell, incorporates a straightening vane in the suction side of refrigerant pumps. And, while cavitation is usually thought of as an issue toward the right (high flow) portion of pump curves, heavily throttled flow can lead to fast recirculating currents that cause cavitation at low flow also. Cornell addresses that side of the curve as well.

We called those vortex breakers.

Yes, commonly designed into piping feeding pumps. Cornell pumps (some of them, anyway) have a fin built into the inlet to prevent pre-rotation.

Thanks to all of you for your comments.

I understood there are more consideration not just NPSH margin while deciding suction pipe size. I am in talk with pump manufacturer and waiting for his reply.

That's not talk. Talk is 2-way and immediate, and involves the use of the telephone. There's room for immediate improvement there.

A detailed drawing would have made this thread much shorter.

They must not have any engineers wherever this is.

This is a waste of time.