Pump Curve Run-out 2

[david6245]

Hello, I do not have much practical experience with pumps and have a question as to what physically happens when a system curve is at a point beyond the rated pumps curve. In other words, what happens when a pump runs out on its curve. Do you simply extrapolate the curve to the lower TDH and higher flow rate? Or can the pump even operate at this point? Would it shut down???

Any clarification would be greatly appreciated.

Thanks,

 

[psafety]

Does it have the installed horsepower capable of such an occurance?

 

[JohnGP]

It's generally not a good place to be. I'm basing my response on centrifugal pumps - as psafety indicated, the power requirement continues to rise, and that may be enough to shut the pump off when the motor overloads, but that's not all.

The pump is not designed to operate continuously off the end of the curve, and doing so can lead to higher vibration, higher temperature rise, high bearing loads, fluid surging, just to name a few effects. Efficiency drops off, and NPSHr increases (which can itself lead to unstable operation).

During commissioning, when analysing results from pump runs, I have found a pump to be off the end of its curve and running quite happily on the extrapolation, but others exhibiting significantly more vibration and with lower flow and head than the extrapolation would predict.

In short, the pump manufacturers don't want you operating in that region, and as indicated above, results can be unpredictable and there are no real benefits.

Just my thoughts,
John

 

[Artisi]

Basically there are 2 main occurences: 1) motor overload, unless of course an oversize motor has been fitted, 2)rising NPSHr and reducing NPSHa which can lead to severe cavitation and subsequent damage to the impeller.

Other problems can be: unstable flow, high axial loads on the impeller / bearings, high levels of vibration and noise all which contribute to shortened pump life.

If there is any chance of the pump "running-out" on its curve you could monitor and alarm power input or fit an orifice plate which will increase head as the flow increases and limit run-out.

 

[BigInch]

At pump runout where flows extend to regions below pump minimum flow, fluid temperature increase also becomes a problem. Any farther than that, gets you in the region of negative flow, reverse flow into the pump and you have to go to the pump curves characteristic of operations outside of the first quadrant. Please don't go there.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[rmw]

Huh??? I need to go back to pump school on that one-it's been some time since I was there. I thought pump run out was where the flow through the pump was at the maximum point. Did you mean minimum head?

But I absolutely do agree with the recommendation "don't go there."

rmw

 

[MartinSr00]

The curve extends. You can extrapolate it to negative heads if you wish. Getting the pump to go there is another problem.

At best efficiency point (BEP) flow, the designer designs the impeller exit blade angles to match passage flow angles. Often inlet blade angles are set optimally at a flow lower than BEP to get a "usable curve" and allow for some fluid pre-rotation in the suction pipe.

As you extrapolate the pump H-Q curve, the passage flow angles will become highly mismatched with blade angles. The problem becomes most acute with regards to suction pressure. NPSH requirements tend to go through the roof when you go significantly above the optimal suction flows. Additionally, cavitation damage is possible. Remember that NPSH and cavitation are two different issues. When there is head falloff because of insufficient NPSH, the cavitation is so advanced that it blocks the flow passages. Damage can occur well below stated NPSH values.

So, to get the pump operating "out on the curve" you may well need a charging pump in series with the pump to boost the suction pressure sufficiently. Increase the capacity of the charging pump enough, and the head can go negative.

If you need to run out on the curve for brief periods of time and have sufficient NPSH, ask the pump manufacturer for advice. I can't think of any case where you would want to order a pump where your normal duty point was way out on the curve. Better to pick another pump.

 

[BigInch]

OK, right, that should be where head goes less than 0, which gives you negative flow anyway.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[stanier]

Check out

http://www.mcnallyinstitute.com

for you free stuff on pumps.

By the way listen to the folks above DONT GO THERE if you want your pump to survive.

 

[MartinSr00]

From my earlier post where I mentioned you could (not a good idea) drive the head negative. When you did this, the flow would be very high in the positive direction. You would be using the running pump as an orifice. If you have a copy of the classic "Centrifugal and Axial Flow Pumps" by Stepanoff, there are some multi-quadrant pump curve examples that would show this.

Don't get me wrong, operating here wouldn't make a lot of sense. But pump companies test the pumps under wierd conditions to assess pump behaviors, etc. during motor trips, surges, etc. When the pumps get very large, such information can become crucial.

 

[SNORGY]

That would venture into uncharted quadrants for me. I'd be lost for sure.

To me, life is much more positive when 0 </= a </= (pi)/2.

You just have to make sure you don't go off on a tangent.

Regards,

SNORGY.

 

[puzzler77]

An optimum H-Q BEP should be sought at a mid-point of conditions. The duty point will shift back and forth along the curve as conditions change. The NPSHa is usually stable in a Water Treatment Plant not fluctuating more than a foot. As demand increases, more treatment is taking place. However, some cities like to fluctuate the level of their water towers to cycle its contents (keep it from becoming stagnant) so I would suggest targeting the mid point of the water tower when calculating the differential head and finding a pump with a BEP.

When the tank is at its lowest point, there will be less head required giving a higher flow rate from the pump. As the tank fills, the head increases decreasing the flow rate.

 

[dabluffrat]

Getting a little too complicated here w/o knowing what type of pump this is. vertical pump for example will go into an up thrust condition before head goes negative...

Limit end of curve run-out performance to 120% - 150% (max) of BEP flow if adequate NPSH is available. anything past this, consider re-rating the pump to a larger impeller dia (if possible) or increasing pump speed (if possible)


Did you know that 76.4% of all statistics are made up...

 

[Pumpone]

Hi,
Operate over BEP to more than 120% you have:
1. More power consumption.
2. More NPSHr with cavitation failure risk.
3. Erosion in high velocity zones, Cut water, suction, discharge vanes zone, etc.
3. More radial load with bearing and shaft failure risk.
4. Global vibration levels increase dramatically.
5. Reliability going down to 30%.
6. Other unstable hydraulics issues.

So it is not recommended.