velocity at pump inlet

[DaniMP]

Hello everybody !
A quick question regarding the velocity in the pump suction pipe :

We have a water pump designed for 100 000 bbld. The water is @ atmospheric pressure and 20 - 80°C in the suction system. I would like to use a 12" pipe for suction, but the velocity would be of 2,5 m/s, and I think that I've read somewhere that it should be below 1,5 m/s.
What would be the risk tfor my pump if the velocity is too high ?

Thanks !
Daniel

 

[Artisi]

The inlet velocity at 2.5m/s is probably too high, but my concern would be the 80C water temp. and its effect on NSPHa / NPSHr - has it been taken into account and its influence on the pump hydraulics.

What diameter is the pump inlet flange and where on the pump curve is the maximum duty point?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

The risk is that you will drop below the NPSHR of the pump or even get close to is and the pump will start to cavitate due to friction losses in your suction pipe.

However there is no such velocity limit as each system is different, so "it depends"... What is probably being referred to is reducing the frictional losses in the pipe, but if ou have lots of metres head and your pipe length is short, maybe it's not a problem. However 1m/sec might be too fast if your pipe is very long.

Water at 80 C has quite a high vapour pressure so you need some more metres from somewhere.

what does it depend on?
Height of liquid level above the pump (+ve) or height it is being lifted (negative)
Length of pipe and fittings
Actual temperature of the water
Elevation above sea level


My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[DaniMP]

Actually I was not very accurate, the water is coming from a tank, and the LSLL is several meters high, so no problem with NPSH (and 80C is design temp, it will be around 40°C).

But just to understand, what is the risk with the velocity ?

 

[DaniMP]

Thank you littleInch !
If the problem is only the pressure drops due to friction then I won't worry, I have only a few dozen meters between the tank and the pump...

 

[1gibson]

What kind of pump, and how are you going to transition from your 12" pipe, to the pump suction flange which will most likely be larger?

2.5 m/s is not that fast (~8 ft/s) so if you follow the rest of the rules (5-10x diameters of straight pipe before the suction, after any fittings like elbows or reducers) then you will be ok with some pump types, should be ok with others.

 

[DaniMP]

Centifugal pump and the flange is also 12" ! (why would it likely be larger ?)

 

[LittleInch]

You should be ok with those sort of figures. the figures you mention are really just a start point guide / rule of thumb if your inlet line is quite long. Pressure drop varies by velocity squared with all other things equal and 2.5m/sec is a bit fast for an inlet line, but should be fine so long as you have the recommended straight lengths coming into the pump nozzle.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[1gibson]

Why would the pump suction flange likely be larger? Because presumably everyone (or at least someone) involved in the pump selection process would have also been aware of recommended suction velocity limits and selected a suitable pump for the application.

If the pump vendor sized it for you and the pump suction flange is 12", then the pump vendor acknowledges that the velocity will not cause problems, and they warranty the pump for those operating conditions.

If the pump was off the shelf and just slapped into this system, then you should provide some more details. There are a lot of different types of centrifugal pumps, is it vertical, horizontal? Some will be fine, some should be fine, some might not be fine.

If you already have the pump with a 12" flange and are going to use it no matter what, then you might be asking the wrong question. Calculate the losses for the 12" pipe and you are either fine, or not fine. Changing to a larger suction pipe will decrease the losses, but it will not affect the suction velocity into the pump. The suction velocity into the pump will be whatever you get with the 12" suction flange.

I take it from the original post that the 2.5 m/sec is the normal flow. A little high but not scary. Will the pump every be operated at higher flow, and if so what is velocity at that higher flow?

 

[DaniMP]

2,5m/s will be for rated flow, half for normal flowrate !
My pump is single stage, single suction, radially split volute type

 

[bimr]

Suction piping should generally be at least one size larger than the suction flange at the pump. The 2.5m/s velocity is on the margin at the high side and should work. However, when you operate at the maximum limits of design practice, you lesser the tolerance in your design.

I see that you are now saying that you will operate at normal conditions with half the flow.

One should carefully determine the operating conditions when installing a pump. The pump operates efficiently at the BEP. When you move off the BEP, the efficiency drops and you may end up wasting substantial amounts of energy over the long term.

If you are operating substantially away from the BEP, then the pump life may also be affected.

 

[DubMac]

If there are solids in the water (sand), then the good rule of thumb for velocity is 5-10 FT/SEC. Below that they can settle out, up around 12 ft/sec and above you can have significant erosion.

 

[BigInch]

How much meters exactly is "several"? For me several means like 3 to 4, maybe 5. "Several meters" is not enough suction head for whole lot of pumps. Pump curves are made assuming the fluid is cold water. Hot water is quite a bit different from cold water, so you should probably add 1.5 meters or so to your NPSHR as well.

I hate Windowz 8!!!!

 

[Artisi]

Come on BigInch, you need to get up to speed with modern engineering language, "is several meters high" and "I have only a few dozen meters between the tank and the pump..." several means less than a lot and a few dozen is a lot less that many dozen but about the same as several.
Of course there is also many others - for example
adequate
enough
plenty
heaps
sufficient
should be ok


It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[HPost]

I can tell you about a very well known UK car manufacturer that has an engine with a water pump inlet velocity of over 7.5m/s.

It also has a volute cast into the block with just about the worst surface finish I have ever seen on a casting.

Needles to say, the pump inlet is 77kPa absolute (cold).

They wonder why they have cavitation damage.