Pump Head and Flow Rate without VFD 1

[p6g2p6]

This question may sound kind of dumb, but this has been confusing me for the last few days.

Here is the scenario. I have a pump system that I've calculated the total pump head for, given a desired flow rate, and I've found a pump that can provide that flow rate at the calculated head. It is crucial that the pump discharges the water at the desired flow rate. What would happen if the actual head of the system is more than what the pump can provide at the given flow rate? Head loss calculations seem to be based off of a lot of approximations, therefore if the calculated head doesn't equal the actual system head, then does that mean that the pump will not provide the desired flow rate?

I hope I didn't make this super confusing.

Thanks in advance.

 

[TugboatEng]

Before VFD's pumps were designed oversized and then had their impellers trimmed to match system requirements. Most manufacturers provide curves for various impeller diameters.

 

[MortenA]

A centrifugal pump has a pump curve, it will always be on this curve. If you plot your system curve in the same plot, then you can find you flow rate - provided that you calculated correctly.

So without control (be it VFD or a control valve) then you cant manipulate your flow rate actively and it will be what it will be - hopefully sufficiently accurate.

But as Tugboat says, you can trim your pump impeller and reduce the flow head and thus flow. This works the same way as reducing the pump speed. You will save some energy although maybe not stricly by the affinity laws due to reduced efficiency.

A few pitfalls: If your upstream or downstream static head changes (e.g. you are emptying a vessel or filling a vessel with a bottom inlet then your flow will also change. If your fluid characteristics changes (e.g. by temperature or composition) then flow rate will also change.

Best regards, Morten



--- Best regards, Morten Andersen

 

[LittleInch]

What type of pump do you have?

If you want a set flow rate regardless of varying head the choose some sort of PD pump.

If you really need a fixed flow within say 2% then unless your discharge head is very fixed you really need some sort of control, be it a control valve or VFD.

Head loss calculations have many variables so yes it can vary if those variables change. The calculations tend to be conservative so often the pump is oversized.


How accurate a flow do you need?


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

 

[1503-44]

The short answer is that flow will reduce, as the pump tries to move the water through it. Eventually the pump's flow will reduce to where it will deliver its maximum head. If the pump's head is still not enough to flow water to your piping outlet, all flow will stop.

When flow stops, the pump continues to run, but does not discharge water into the piping. That will cause the pump's expended energy to heat the water and the pump. The higher temperatures may lead to pump damage.

 

[pierreick]

Hi,
The result can be seen at the intersection of the pump curve and the system curve !
Note : Check the impeller of the pump , you may want to change it to a bigger one if not the case . For that you need to review the set of curves given by the manufacturer or call him .
Good luck.
Pierre

 

[Artisi]

If the pump selection was based on a calculated head /flow at x speed, y impeller diameter and z suction pressure and the installed operating head is higher, the flow rate will be less - supply more data for a useful (maybe) answer.

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.)

 

[1503-44]

The problem is that, if system head is higher than pump's maximum shutoff head, the two curves do not intersect.

 

[LittleInch]

as usual we have about 10% of the data [sigh]

We don't know:
type of pump
flow rate needed
head calculated
pump curve
system curve or range of head for the same flow rate
how accurate the flowrate needs to be. Is more flow just as bad a s less flow?

If it is "Crucial" that the flow rate stays within a certain unknown band, then you need some form of control on flow - be it a control valve ( the easiest and simplest) or a VFD.

I await answers....


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

 

[EdStainless]

If you look on the pump curve you should see upper and lower recommended flow limits.
Depending on the style of pump they may be +/-20% from BEP.
Now look at your assumptions and think how different reality may be.
Do you still fall in that range? If so you are fine.
If there is a chance that you will need to run at higher head and lower flow either select a different pump or instal a recirculation line with a control valve in it to start with.
Don't run off the right hand (high flow) side of the curve.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[p6g2p6]

Thanks everyone for the replies - all are very helpful, and I greatly appreciate everyone taking the time to reply.

I've reached out to local pump supplier and they provided me with a pump curve for the impeller size that is required for the selected flow rate (see below)

The flow rate that I need is 120 gal/min (+/- 5%). If I want to avoid using a VFD, would it be best to slightly oversize the pump and then use a flow control valve to ensure that the flow rate remains constant at around 120 gal/min?

Thanks

 

[1503-44]

You mean oversize the impeller at 7.5 or 8" or so.
The pump is already oversized, judging by its best efficiency point.
Yes. Don't even think of vfd control for such a small variation of flow rate.

A vfd would also reduce head as the rpm and flow slowed down. They generally don't work best for constant speed and head applications. Flow variations should be greater than 30% to start thinking about vfd.

 

[LittleInch]

For what you are trying to do this is a very poor choice of pump.

The duty point is too low for this unit - poor efficiency and you are in a very flat portion of the curve. Very small changes in differential head make a big swing in flow.

~your pressures are very low as well so a control valve will need quite a big increase (10-15 ft?) to work properly.

Why don't you use a PD pump? you will get your +/- 5% if you size it right and then you can more or less forget about changes in frictional head loss.

You need also to work out your system curve - flow vs head.

Also remember a pump is differential head so any changes in your inlet head makes a difference to the outlet pressure.

We now have about 30% of the information....


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

 

[EdStainless]

Look for a pump with a steeper flow curve, it makes control easier.
With this small of a pump I wouldn't worry about power consumption, but running off of BEP by more than 20-25% can really shorten the pumps life.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[p6g2p6]

I didn't select this pump - I provided the desired flow rate and head to a local pump supplier and this is the pump that they provided a quote for. The main reason why I requested a quote was to get a rough cost estimate for the project, as I have an Interim Design report coming up soon (this is for my senior year project).

LittleInch, I haven't considered a positive displacement pump yet, however I am open to any ideas, therefore I will certainly look into that option. Thank you for the suggestion.

EdStainless, thank you for the advice - I will continue to look around for other pumps that better fit the flow rate and head.

 

[danschwind]

IMO, there is nothing wrong in controlling a pump this size with a VFD. This is a very small pump and VFDs are affordable nowadays.

As others have pointed out, there are probably a couple different ways to reach the same goal, each one with advantages and disadvantages. If you want better advice, then offer us more details of the overall system.



Daniel
Rio de Janeiro - Brazil

 

[p6g2p6]

Okay I will try to explain the system as best I can.

I am trying to design a pumping system that pumps discharged cooling water from 5 water-cooled compressors to a mixing tank that is 15' above the compressors. Cooling water exits compressors at 30 gal/min, therefore there the total flow rate is 5 x 30 = 150 gal/min. The pressure of the discharged cooling water is unknown as of right now, however it will be determined shortly and I'll be able to include the pressure on the suction side in my calculations (so far my calculations have assumed that the pressure of the cooling water on the on the discharge side of the pump is zero, which is conservative). The pump needs to be able to provide 120 gal/min to the tank. This is less than the flow rate of the cooling water exiting the compressors, therefore I'm going to have a 3 way control valve that lets 120 gal/min through to the intake of the pump and discharges the excess flow into a drainpipe.

Will a positive displacement pump work if there is already a relatively high flow rate on the suction side of the pump?

 

[Artisi]

????? Discharge from the compressors is 150gpm, return to the tank is 120gpm, where does the shortfall of 30gpm of cooling water come from???

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.)

 

[pierreick]

Hi,
A sketch with figures will save words and will be understandable for everyone .
Note : to me the outlet of the compressors should go to a tank equipped with an overflow directed to the drainage system for the surplus ( no need for 3 way valves ) .From the bottom of this tank the outlet is connected to your pump.

Pierre

 

[FacEngrPE]

Don't over think this system.
Compressors are not very picky about flow rate as long as it is equal to or more than the minimum specified. You do need to have some method of balancing the flow through the 3 compressors, but that should be adjusted to minimize the balancing valve head loss. Most compressors like cold water, when oil cooling is present usually that part of the system has a temperature control valve provided, to maintain the minimum allowed oil temperature. Hopefully the O&M instructions explain this.

Flat pump curves are ideal when you are pumping from a pond and want a outlet pressure that does not change much with flow changes. You can design your system with this kind of pump, but some of your components choices will be different than when using a pump with a steep curve or a PD pump. Previous comments about this being a not very efficient pump selection for a flat curve pump are correct. Tell your pump supplier to do better.

I would avoid use of a VFD on a compressor cooling loop, it is unnecessary, and adds to the count of parts that need to be maintained. There is some merit to VFD on cooling tower fans if they are large enough at 150 gpm your cooling tower might be large enough to investigate that. The tower fan speed in this case is adjusted to maintain water loop temperature between the required limits.