Recommended Pump Suction and Discharge Side Velocities

[Pavan Kumar]

Hi All,

I wanted to get hold of a standard reference which mentions the recommended velocities on pump suction and discharge sides for a centrifugal pump. I understand the on the pump suction side NPSHA governs the velocity but a general guideline and reference would be very helpful. One of the online reference, link copied below, mentions the recommended pump suction and discharge velocities as 3-6 ft/sec and 9-12 ft/sec respectively. Norsok standard P-001 Rev 3 mentions the maximum velocities which look more erosional velocities. I am looking for more like a company standard or some good engineering practice document that mentions this.

https://www.pumpfundamentals.com/ce...e will initially,energy cost will be greater.

Thanks and Regards,
Pavan Kumar

 

[LittleInch]

You've found one. What more do you want?

Each pump system will be a little different but for liquids those velocities are a good starting point.

I think you're looking for something which doesn't exist. There are many many "guides" and recommendations out there but no one "bible". Just like their isn't one pump vendor or one design company.

Why do you want this?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Those velocities are very good to use as guide lines. I have found that website very reliable.

Those velocities are not for erosion control. They would be based on providing manageable NPSHA and avoiding waterhammer and discharge pipe sizes that are unreasonably small to be practical and economic. One could easily double or more the discharge velocities, but you had better know how potential and severe waterhammer pressures might become and if your discharge pipe pressure drop might be too great to run economically. It makes no sense to buy a pump only to burn too much of its discharge pressure on simple bad pipe design.

 

[Pavan Kumar]

Hi 1503-44 and Littleinch,

I sized a pump and the piping system recently with a rated flow rate of 75 US gpm. The pump suction line size is 3" sch 40 while the discharge line size is 2" Sch 40. The pump is a 2"X1"-10" which is 2" suction side flange, 1" on the discharge and 9" impeller. The velocities and pressure drop in pump suction and discharge lines are 3.25 ft/sec, 0.6 psi/100ft and 7.17 ft/sec, 6.6 psi/100ft respectively. I felt getting the discharge side pressure drop to 3 or 4 psi/100ft would be good but that will result in 2.5" Sch 40 pipe. With 3" Sch 40 pipe the DP is 1 psi/100ft. The velocities in 2.5" and 3" pipe are 5 and 3.25ft/sec. I do not want to use 2.5" pipe as it is not a standard pipe size and I do not want to use 3" pipe as the cost of 300 ft pipe would be very expensive. Power required at duty point for 2",2.5" and 3" discharge pipe are 2.54,1.76 and 1.36 HP respectively. The cost of electricity for 2.54HP,1.76 and 1.36 HP power requirements are $2074, 1437 and 1110 per year respectively. I just want to make sure that I am not sizing a line that will introduce unreasonable permanent pumping cost due to smaller than required line size. Besides I want a reference to justify my line size. Mentioning a standard engineering guideline or reference would help this.

Thanks and Regards,
Pavan Kumar.

 

[Nutzman]

Hi Pavan,

Please clarify the following in your original post:
"on the pump suction side NPSHA governs the velocity...."
I fail to see how NPSHa governs velocity. You could have a suction tank on a hill and the pump down the hill, with lots of suction head, a high suction line velocity and wreck the pump.
"maximum velocities which look more erosional velocities...."
Erosional velocities? Erosion occurs as a result of solids been pumped? Are there solids in the fluid?

A simple means of determining suction and discharge velocities is to look at the operating point on the pump curve.
Take the pump inlet diameter and the do the pipeline velocity calculation based on flow and line size. Then repeat for the diameter of the discharge size.

When designing suction pipe diameter, the "rule of thumb" is to increase the suction pipe one (pipe)size. By a standard pipe size. Eg. if the pump inlet is 150mm then fit a 200mm suction pipe. Using an eccentric reducer (flat on top)to take the pipe from 200mm (pipe size) to 150mm Pump suction.

Then it comes down to the economics of cost per meter of discharge pipe, versus power draw, motor size and cost of switch gear.

 

[LittleInch]

Looks to me like you're doing a good job there where, although small scale, you are balancing lower pressure drop and hence lower power usage, with higher one off capital cost of the 3" pipe. It's called design.

That's all you need to do - you don't need any "standard engineering guideline" because

a) it doesn't exist and
b) it is only a guide anyway to use as a starting point.
c) every system and client is different

It is very common for the pump discharge to have a smaller diameter than what is economic for a long run of pipe, but it's cheaper and better for the pump to do this as they don't know if your discharge line is 10ft long or 1000ft long.

It is actually fairly rare for anyone to actually build a bigger pipe and save on power costs, but it does happen, usually only if the client is left paying both CAPEX and OPEX....

If your pump ran for 10 years then difference between 2" and 3" is $10,000. I'm finding it difficult to believe that 300ft of 3" pipe is that much more than 2", but do the maths yourself.

At these sorts of sizes you get a massive jump in things from one size to another, but the principle remains for bigger pipes.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Pavan Kumar]

Hi Nutzman,

I should have stated that Pump NPSHA should be kept in mind while sizing the pump suction line while keeping suction line velocity should be reasonable(that is above the minimum velocity). What that reasonable velocity is what I am asking in this thread.

With regards to the maximum velocities mentioned in Norsok standard, I felt the maximum velocities seemed to be very high, close to the erosional velocity called using the formula Ve = [C / SQRT(RHO)] where C = 100-160 and RHO= fluid density in lb/ft3.

Agreed high velocity does not cause erosion unless there are vapor bubbles are solids present in the fluid. My intention is to get reasonable velocities in pump suction and discharge sides.

Thanks and Regards,
Pavan Kumar

 

[Pavan Kumar]

Hi LittleInch,

The higher cost of 3" pipe will pay for it itself from the saved power cost on the long run. But you see the velocity in 3" pipe is only 3.25 ft/sec, the recommended velocity is between 9-12 ft/sec per Norsok standard. Some other sources say it is 6-9 ft/sec. I guess as long as the velocity is more than minimum velocity limit which is 0.8 m/sec ( 2.62 ft/sec) it is ok. Having 3" pipe on the suction and discharge side does not look normal though. Usually the fluid velocity is higher on the pump discharge side.

Thanks and Regards,
Pavan Kumar

 

[LittleInch]

The velocities you wrote are "typical", not "recommended". Big difference.

I don't know where 0.8m/sec comes from either?

When you include OPEX costs the optimum pipe size usually goes up a pipe size or two.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Pavan Kumar]

Hi LittleInch,

The pump discharge side velocities I mentioned are typical. What are the recommended velocities then?.

The minimum velocity of 0.8 m/sec is from Norsok Standard P-001 Rev 3 Page 13/29.

Yes if I consider the pumping cost then the discharge pipe size increases from 2" to a 3", but it appears it is usual to choose the 2" size even though 3" pipe size will be economical in the long run. In all cases the CAPEX and OPEX are both paid by the client isn't it. In my case I am designing the system for the plant that I work for, so I need to be all the more careful. It is like I am spending my own money.

Thanks and Regards,
Pavan Kumar

 

[LittleInch]

That 0.8m/sec only refers to liquid with sand or silt.

A lot of times people are only interested in CAPEX, therfore that's why smaller pipes are more common.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[TiCl4]

Pavan,

What is controlling pump flow? Do you desire a higher flow than the 2” can give you? If not (i.e. if you have a flow control valve downstream that always controls to 75 gpm), then there is no point in increasing line size - you will just take the dP across the valve rather than the piping.

To put it another way, if you will operate the pump at 75 gpm, you will use the same power no matter the downstream pipe size. If your 75 gpm setpoint will not change, then the savings for increasing pipe size will only come if the pump impeller size or pump speed can be reduced to provide 75 gpm at a lower head pressure.

Be aware that such a large difference in line losses (1 vs 6 psi/100 ft) can significantly affect a flow control valve sizing and operation, if that is what is controlling pump flow. Lowering pump head may also not be feasible for other reasons (like a high static head that must be overcome).

 

[Pavan Kumar]

HI TiCl4,

There is no controlling element that will control the 75 USgpm flow rate. I am getting the desired 75 gpm flow at the duty point. I probably may have to throttle the pump discharge valve to make the system curve and pump curve intersect at 75 gpm flow rate.

Thanks and Regards,
Pavan Kumar.

 

[Pavan Kumar]

Hi All,

With regards the suction side line sizing of the pump, I found that having low fluid velocity will allow the solids in the fluid to settle down. In my case the fluid being pumped is a thin slurry with 12% solids in it. I calculated the terminal settling velocity as 0.10 ft/sec using the formula for Terminal settling velocity. Attached is the excel spreadsheet. I wanted to know if this number makes sense. My suction side line size depends on this criteria also.

Thanks and Regards,
Pavan Kumar

 

[georgeverghese]

Given that you've got solids in this stream, a smaller ID line would give you better turndown on flowrate in order to avoid nuisance blockage of this line at low operating flows. Blockage at low flow can lead to production loss which may most likely far outweigh any power savings with the larger ID line.
Install rodding points at all elbows to enable mechanical cleaning and avoid low points in the piping.

 

[Artisi]

Finally after many days and having being asked previously about solids in the stream we are now advise on that small insignificant point, being told initially would have saved many people a lot of time and effort.
All you need to do now is give some detail on the solids without people having to drag the information out bit by bit.

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.)

 

[Pavan Kumar]

Hi georgeverghese,

We propose to install air/steam blow out points in the pump suction and discharge lines to avoid plugging. My issue is to ensure that the velocity in the suction line to avoid settling. For this purpose I calculated the terminal velocity shown in the spreadsheet attached above. I want to make sure my calculation is correct.

Thanks and Regards,
Pavan Kumar

 

[Pavan Kumar]

Hi Artisi,

The fluid being pumped is Waste water with 12% solids. Following is thee data regarding the fluid and solids:

Fluid : Waste Water
Temperature : 65 Deg C
Density : 66.646 lb/ft3
Viscosity : 10 cP

Particle Data
Particle Diameter = 1000 Microns
Particle density = 102.96 lb/ft3

Based on my calculation using Stokes Law I got a terminal velocity as 0.10 ft/sec. Please see calculation attached. Does 0.10 ft/sec make sense is my question.

I have used 3" Sch 10 in the pump suction line which gives ma velocity of 2.88 ft/sec for the pump flow rate of 75 US gpm. I want to make sure this velocity is high enough to prevent the solids from settling. With 3" Sch 10 pipe NPSHA is 12 ft. I cannot use 2" Sch 10 pipe as the NPSHA falls down to 6 ft. NPSHR is also 6 ft.

Thanks and Regards,
Pavan Kumar

 

[TiCl4]

Pavan,

There is no controlling element that will control the 75 USgpm flow rate.

I probably may have to throttle the pump discharge valve to make the system curve and pump curve intersect at 75 gpm flow rate.

Those statements are directly contradictory. If your second statement is true, the pump discharge valve is the controlling element. Regardless, my main point was that you will not get any savings on power usage by going to a large pipe simply because you'd have to then throttle a valve down somewhere - your pump will always run at the duty point of 75 gpm. You'd need to both increase the pipe size and decrease the pump rpms or impeller size in order to reduce pump power draw.

Secondly, your calculation of settling velocity may not be correct. Stokes law is accurate up to a Reynold's number of 0.3.* I haven't run the numbers, but my eyeball judges that you will have a Re that is higher than 0.3. Also, you have a high enough solids content that hindered settling will affect your settling velocity. At 12% solids, your hindered settling velocity will likely be anywhere from 60-70% of the calculated normal settling velocity.*

*See Perry's Chemical Engineers' Handbook, sections 5-61 to 5-64.


Finally, if all you want is a means to keep the material suspended, I refer you to Perry's again, sections 5-46 and 5-47, where they discuss settling velocities, particularly equation 5-151. This equation is applicable up to 1 mm particle diameter. You did not give a particle size distribution, but just said 1000 um (or 1 mm), so this equation is on the borderline of applicability. 80% of your particles must be 1mm or less for this to apply.

With your given information, I get a minimum transport velocity (horizontal pipes) to avoid settling of:

2" NPS: 0.63 ft/sec, or ~7 gpm
3" NPS: 0.83 ft/sec, or ~20 gpm

Given you will be flowing 78 gpm, you will be at least 3x the recommended settling velocity of this equation even for a 3" pipe. However, if your particle size were only a little bigger (2mm), then equation 5-153 would be applicable (it's for 2mm and larger particles) and required setting velocity would jump to:

2" NPS: 3.7 ft/sec, or ~38 gpm
3" NPS: 4.5 ft/sec, or ~104 gpm

I don't know how much control you have over your particle size, but a 2" pipe should give ample coverage for a larger range of particle sizes at 78 gpm than a 3" pipe. If you install a 3" pipe, then some process upset that creates large particles may cause plugging issues. It appears that a 2" pipe would be a safer option from a settling perspective, but 3" will be just fine as long as your particle size remains at your current stated particle size.

 

[LittleInch]

I probably may have to throttle the pump discharge valve to make the system curve and pump curve intersect at 75 gpm flow rate.

Then forget all your previous calculations about power.
You match the pump curve to the system curve unless you've already bought your pump...

Otherwise any power savings don't exist.

And 0.8 to 1 m/sec is a good number to use as a starting point if you've got solids flow going on.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.