Pump Curve Has a Maximum Head 2

[bulkhandling]

This is an abnormal curve. When the flow increases from zero flow, the head gets bigger till to a maximum point and then the head get lower like a normal curve. I know this kind of curve is not good but do not know what may happen or what will be the impact in the operation.
Need your help.

 

[friartuck]

What sort of pump is it?

Pump curves and fan curves are similar, they only pump different fluids. Fan curves are like that you described, however what manufacturers usually do is to publish the curve that is either the most efficient or where stable flow is probable. (like fan curves)

You will probably find that the curve is pretty normal, just dont use the 'bad part' of it.

The pump manufacturer (if they are any good), will have the power absorbed curves and efficiency curves for your pump. If they can't give these to you, you might be dealing with a low budget pump manufacturer who doesn't have the required test facilities.

Generally, use the RHS of the curve and dont go above about 75% up it. This will give you reasonable operating conditions without causing problems.

go to the mcnally institute for further info.

http://www.mcnallyinstitute.com

Friar Tuck of Sherwood

 

[TZellers]

Sounds like what is known as a drooping curve. The peak head is produced just just of shut-off. This can be caused by several things, but revolve around the pump design itself. If you are operating in the region where 2 flows exist for the same head, you can get into oscillation issues and flow control problems.

Usually too may vanes cause a drooping curve. There are some things that can be done to correct or offset the droop: put a 15 degree or higher bias on the trim of the impeller (i.e. trimming is performed so that it is tapered at 15 or greater degrees from front to back, so that the rear shround is smaller than the front). Also, scalloping the impeller (removing material between the vanes at the perophery of the impeller) but will effect efficieny by about 3-5 points.

 

[quark]

That is called an unstable curve. If you allow your pump to run at any other point except the maximum head, there will be hunting (i.e pump operates at the two flow points corresponding to the head on either side of the peak point. I have yet to find out the actual advantage of such a pump. there have been many discussions on this topic in this forum but not a concrete idea evolved so far.

Regards,

 

[25362]

Speaking of a single pump with a drooping characteristic curve, as long as the system curve at shutoff (static head) is lower than the pump's shutoff point, there wouldn't be instability, and the curves would intersect just once.

However, when, for example, the mains frequency drops, from, say, 50 Hz to 47.5 Hz, the system curve may cut the pump's curve at two points bringing about unstable operation. And, as you know, stable characteristics are a basic necessity for the successful automatic control of centrifugal pumps.

 

[bingopin]

This type (shape) of pump curve is not that unusual. You see it on high head pumps such as Sundyne or Sunflo. A characteristic of this type of curve is intolerance to low flow. For the reasons stated above, low flows or operating points "back" on the curve can make pump operation unstable. No problems if you keep the operating point "out" where the curve has some slope to it.

 

[bulkhandling]

Thanks for all your comments. In my application, I know it will be ok - pipe length is about 1000ft, friction loss constitutes about 1/3 of the total head. But I do not like this curve and just want to give a reason to deny it.

 

[quark]

It's better to deny it always rather than having to change the bearings frequently because of thrust loads due to hunting.

 

[smckennz]

There is absolutely nothing wrong with the curve you describe. Just make sure that your operating point is reasonably lower than the shutoff head and you will not have a problem.

I have occasionally operated pumps in the "grey" area above the shutoff head; more for curiosity than to do anything useful. Never had a surge problem (unlike when I tried it with a squirrel cage fan), although the friction component was high in relation to the lift. You will normally find that the peak pump curve head is well to the left of the BEP, and you have no business trying to run a pump there anyway.

Cheers

Steve

 

[TZellers]

I should also note that it is not uncommon for some of the pump manufacturers to publish...er...um...art instead of science. In other words, some marketing types have taken the liberty to eliminate the droop graphically, not in actuality.

Another piece of typicaly artwork on pump curves is the NPSHr curve. Generally speaking an accurate NPSHr curve will rise as you approach shut-off, but again most companies draw it horizontal at zero flow.

Be careful what you reject because of a curve. Is one telling the truth and the other not? Put in the spec to have a non-drooping curve and get a performance test.

 

[ccfowler]

There is nothing inherently wrong with the type of pump curve that you describe so long as it is compatible with your system's characteristics AND there is only one pump in operation. If pumps with such curves are in parallel operation, then operation can indeed become painfully "interesting."

 

[friartuck]

I used to work for a fan manufacturer and the curves described are quite normal. Most manufacturers publish the stable part of the curves and 'grey out' the less useful part. However this doesn't mean you cant use it, its just unstable,
Let me explain, at unstable points on the curve, the flow can become pulsing and cause vibration etc. but the fluid is still moves. also at some point on the curve, there can be 2 operating points and the pump/fan will hunt between curve points.
If the fan is used for emergency purposes, i.e. for only intermitent or one off conditions like fire extract, then its not really a problem since the fan is a one off use anyway.
Operating something long tern obviously isn't recommended.

There are pro's and con's for publishibg the whole curve. If you grey out the curve, then the user is not tempted to use the greyed out section. (But some people assume that the curve simply carries on up. (Fatal mistake).
If the manufacturers shows the whole curve, there is the risk that the eselctor can try and use the curve at an inappropriate condition.

Try to use the curve no more than 65-75% from the base to keep well in and create stable operating conditions. People quite often clip the underside of the curve and get away with it, but its not worth the risk.


Friar Tuck of Sherwood

 

[TZellers]

friartuck, I like the idea of greying out the undesirable parts of the curve. Typically, pump manufacturers will draw the curve from shut-off to run-out (approx. 125% of BEP). Then they list (or should) a MCSF (minimum continuous safe flow). Your example is cleaner. It does depend however, on your customer base. In the consumer market, people don't understand why the curve doesn't extend to the abscissa. Then you get stupid calls.....ooops off subject

 

[scalleke]

The pump curve described is normal in for instance Sundyne pumps.

It all depends. It is best not to operate the pump above shutoff head. However if you have to make sure the intersection angle between the system curve and the pump curve is as small as possible.

Pumps can be operated in this area as long as the system inertia is high enough to prevent the flow from shifting.

I do not see how you can refuse the pump curve if a rizing curve to shutoff was not specified. Is per API 610 not only the operating point on the curve guaranteed?

By the way you can always install a discharge orrifice to make the curve continuously rizing to shutoff and if you do that you will simply have to pay the wasted energy.

Best Regards.

Scalleke

 

[PUMPDESIGNER]

Late getting in here, have been absorbed in business stuff.

Drooping curves are not inherently bad, but they are tough for control purposes in many but not all applications.

Stepanoff points out that the most efficient design of any scroll pump ALWAYS has a droop near shut off. Many pump manufacturers sacrifice a little efficiency in order to eliminate that droop.

Some manufacturers are not concerned and many or most of their pumps have the droop, Berkeley is one example of a company where many of their pumps in a certain range have droop.

At least one company we know of hides the droop by providing the performance curve starting at 40 gpm.

The systems we have seen where the droop is not a control problem are where the system is very simple, they start the pump, run at BEP or higher, and never try to really control the pump, just let'er rip. No pressure controls, no safeties, nothing, just start 'er up and let 'er run. Ag is the place for those pumps.

For those of us who must control the pump properly droop is a disaster.

PUMPDESIGNER

 

[jonelv]

There should be no problem with a drooping head curve providing you are operating in the stable part of the curve - as per other comments. Drooping head curves are often desirable because of the greater efficiency

Where you must be careful is if you wish to operate the pump in parallel with another. Here you may get problems due to surging and pressure fluctuations if the pumps don't share the load or you may find the assist pump cannot even start because the head at zero flow is lower than the line pressure.

So I would not operate drooping head pumps in parallel or where significant turn down / control needed but otherwise go for it.

jonelv

 

[madhankumars]

I have got an assignment in which I've gotta test a centrifugal pump.
its a pedal driven(manually operated pump).
I determined discharge head values alone.
wil it do or should i consider the suction too(actually we conducted tests by suctionfrom a barrel at ground level)
hope someone will help out

 

[Artisi]

start a new thread for this and add some more detail if you want help

Naresuan University
Phitsanulok
Thailand