8 replies to this topic

[go-fish]

Please can anyone advise what criteria is used by centrifugal pump manufacturers to determine the pump suction and discharge nozzle sizes. For the connected piping, engineering contractors have a criteria on the upper limit of flow velocity which can be applied to the suction and discharge piping. I have a situation where the pump nozzles is up to 5 times smaller than the connected piping and the manufacturer is stating pump nozzles sizes cannot be increased as they are cast as one with the casing. But I presume the manufacturers may be following some velocity criteria as otherwise velocities > 6 m/s in the 5D straight length section as required by API RP 686. 

[Bobby Strain]

The build pumps most likely from API standards. I presume that they supplied performance data including NPSHR. That's what is most important.

 

Bobby

[Pilesar]

I am not sure how to interpret your question. If the pump is rated for the flow and head required for the fluid conditions, then its nozzles are correctly sized. One of the criteria is to have a uniform flow pattern at the impeller to avoid excessive vibration which would stress the bearings. Suction piping should generally be larger than the pump inlet anyway. As long as there is enough pipe length for uniform (non-swirly) flow then I don't know if there is a limit on how large the suction pipe could be relative to the pump flange. You can likely choose a larger pump and run it at a lower speed to get the same performance with some additional reliability.

[go-fish]

My concern is that pump suction is 10 inch and there is a 5D straight length requirement before adding a reducer (per API RP 686) to match the connected suction piping of 24 inch. The suction piping is sized as 24 inch to limit the velocity < 1.5 m/s per contractor design guides. In the 5D straight length of 10 inch, the velocity is 5.9 m/h and rho x v2 is 34,000 kgm/ms2. 

 

Should I be concerned about flow induced vibration and piping erosion in this small ~1.5 m long section of piping? It is a cooling water pump and in continuous service.

[Pilesar]

I see what you mean in API RP 686 (3.1.2.6 in my copy.) We must be interpreting that wrongly as it is extremely common to put a one-pipe-size reducer directly to the inlet flange to accommodate a larger suction line. Perhaps someone else will supply a clear authoritative reference. Here is one guide that might be useful:  https://www.pumpfund...on piping 1.pdf

[Bobby Strain]

Have you consulted the manufacturer?

 

Bobby

[breizh]

Hi,

Rules from specialists:

https://blog.craneen...gal-pump-piping

https://pdhonline.co...m134content.pdf

Breizh 

[go-fish]

I see what you mean in API RP 686 (3.1.2.6 in my copy.) We must be interpreting that wrongly as it is extremely common to put a one-pipe-size reducer directly to the inlet flange to accommodate a larger suction line. Perhaps someone else will supply a clear authoritative reference. Here is one guide that might be useful:  https://www.pumpfund...on piping 1.pdf

 

 

Yes the concern is regarding Section 3.1.2.6 which states The pump suction line shall have a minimum straight run of five pipe diameters between the pump suction flange and the first elbow, tee, valve, reducer, permanent strainer or other obstruction. The straight run shall be of the same line size as the pump suction nozzle.

 

There has been an ambiguity around this requirement. Although API 686 is Recommended Practices and not a Standard but it states the requirement with a "shall".  I know from previous experience that some vendors recommend the reducer directly at the pump nozzle and straight length requirement does not include the reducer. Normally, it is not a concern but in this case pump nozzle is 10 inch and piping is 24 inch, so there is a big different in velocity depending upon where the reducer is placed. 

[Pilesar]

24 inch suction line is too large a requirement for a 10 inch pump suction. Something seems not right. If the suction line is designed purely for a velocity criteria, I would look for a way to justify bending that rule.