centrifugal pump amp overload and cavitation on high end 7

[drdoom]

Hi Folks,
I am hoping someone can give me a "down to earth" answer regarding a problem i am experiencing in a newly constructed water treatment facility.
I have two seperate applications that are giving me problems.
A) Two 3 hp centrifugal recycle water pumps; operating in conjunction with a VFD for flow control(one as lead the other as standby) are used to lift waste water from a membrane filter unit to a membrane filter dedicated to treating this water. The water is "dumped" into a feed tank from which the skid mounted pump then provides the energy to put the water through the filter. The recycle pumps can serve double duty in that in the event when the dedicated filter is down, the water in the recycle tank can be pumped directly to the waste tank for disposal.
These pumps therefore, can see two different head conditions, in the recycle mode +/- 19.5 ft and in the waste mode +/- 10 ft.
B) Three centrifugal raw water pumps operating in conjunction with VFD for flow control. There can be any configuration of pumps running , from one to all three, depending on demand. These pumps take water from the reservoir , under flooded suction of an average head of 19 feet, then pump it to the feed tanks on the filter skids. The centerline elev. of the pump inlet is 690.00, however the reservoir water elev. average is 707.00 and the feed tank inlet centerline is 749.00. i calculate the total head to be 749.00 - 707.00 = 42 feet give or take and the pipe is 12" X 900 feet.
Ok now the problem; in both of these cases when a pump is operated at 100% of speed the amperage draw exceeds the maximum allowed by nameplate specs. In the case of the recycle pumps (maximum output = 300 GPM @ 20FT) the rated amps are 3.71 and the 100% demand is 4.3.
In the case of the raw water pumps (maximum output = 900 GPM @91 FT of head)the rated amp draw is 44 amp and the 100% demand is 46+. Also i want to mention in the case of these pumps i have to limit the upper limit to 90% of speed or the pump goes into very noticable cavitation.
In both cases, when it becomes necessary to operate the pumps at 100% of speed or slightly less the excessive amp draw causes the VFD to go into overload and shut down. The design parameters say it should work but the reality is they don't. Has anyone any experience with this situation? what is happening and why do the raw water pumps show excessive amp draw (a condition i would normally attribute to a pump working too hard) when it is cavitating (a condition i would normally attribute to not enough suply or "whirlpooling" niether of which is occuring and since the pump is not moving as much water as it should I would expect less amp draw).
I know this is a looooong question but i would really appreciate some suggestions.
thanks

 

[BigInch]

What's the piping configuration look like just in front of the pump's suction? Have any turns there? Do you have 5 to 10 diameters of straight pipe before the suction flanges?

http://virtualpipeline.spaces.msn.com

 

[JJPellin]

You should verify your power supply to the motors, as well. If the voltage is lower than the original design specified, the amperage will be higher in proportion. Make sure you have full voltage.

Johnny Pellin

 

[drdoom]

To BigInch,
the piping configuration is tight and there is certainly not even 5 diameters of straightr pipe prior to the inlet to the recycle pumps and the raw water pumps are manifoled together so the piping makes a 90 degree turn into the pumps. The recycle pumps essentially do the same and the pipe size for these is 4".
To JJPellin,
i have no problem with voltage, as a matter of fact we typically will run 490+ on a 480 three phase circuit.
thanks guys.

 

[TenPenny]

I wonder who spec'd the motors? Sounds to me that the motors may have been sized for 'normal' operation, with no consideration for what happens at full speed.

 

[BigInch]

I would suspect it is "prerotation" at high velocities that is causing the cavitation at full speed.

10p, wouldn't the motor normally be sized for flows at the rated pump speed, which I would also assume is the 100% speed marker on the VFD. Didn't check the power reqmts, but I get the feeling that its not a motor size problem.

http://virtualpipeline.spaces.msn.com

 

[Artisi]

What are the motor sizes fitted to these pump units, firstly you need to establish if the motor sizes are in fact correct for the duty.

On the raw water pumps, it is unlikely to be cavitation in the true sense but extremely poor inlet conditions with very turbulent flow into the impeller eye causing the noise, however, whether cavitation of poor entry into the pump you are stuck with the problem due to the poor pipe configuration.

 

[hyposmurf]

If at any case the system head is under the pump head , the pump will handle more flow than espected , hence the power will increase.
I had seen a lot of burn motors when centrifugal or fan work on free out.
Cavitation is also a concequence of it , as flow is incremented.

Did you have a gauge at in and out of pump , what is the difference??

 

[ccfowler]

drdoom,

I suspect that no small part of your problem may be an unfortunate dependence upon VFD's to serve as a substitute for a truly careful, thorough, and understanding selection of the pumps and design of the system as a whole. This is an unfortunately frequent problem. VFD's can be very beneficial, but they can't make up for a misapplied pump. The short suction lines probably hurt even more if the pumps are being forced to work well away from their BEP's especially at greater flows since the NPSHa's will almost certainly be getting smaller.

The key here is that pumps ALWAYS work on their performance curves. Study your pump's curves vs. the curves of the connected systems. Wherever these head vs. flow curves cross under any operating condition, that is where they will be working. The use of VFD's to cover a range of system configurations suggests that the pumps may be operating well away from their BEP's much of the time.

Look very closely at the NPSHr curves for the pumps. These show the conditions where the pump's performance has already suffered a measurable loss of performance due to cavitation. (There are several standards for the NPSHr curves, but by far, the most common ones are for either 1% or 3% performance loss.) When the flow rates at the pumps are greater than the BEP, the pump efficiencies (and flow patterns within the pump) get progressively worse. It is no surprise that these are troublesome conditions for your pumps. The "cavitation free NPSHr" is usually a multiple of the published NPSHr curve. This multiplier can range from about 1.5 to as much as about 20. As a general rule, if NPSHa is anywhere near the published NPSHr, the application is likely to be troublesome. It may work, but such an application should be avoided if possible. (In general, extensive pump operation beyond the flow rate range of the published NPSHr curve should be avoided.) If you are stuck with such an application, the operation and maintenance burdens have to be balanced against the costs and difficulties of correcting the situation. It is not unusual to simply live with such a mess, at least for a while, because the cure is worse than the problem.

For these relatively small pumps, it is unlikely that you will have test bed performance curves for each pump. When you are looking at the published curves for the particular pump model, the actual performance of any one pump can be expected to be within about 10% of the published performance data. Most pumps will perform reasonably close to the published data, but you may want to check a pump that seems to be performing particularly poorly.

You may want to consider contacting the pump manufacturer(s). They can be most helpful, especially when the pump user is both knowledgeable and reasonable. They don’t want their pumps to be performing poorly anywhere, and they are almost always able to provide very useful application information based on related experiences.

 

[BigInch]

ccfowler,

As much as I would like to agree that its another case of misapplied VFD, I don't think so. Two head requirements for the same pump under different flows could lend itself to a valid solution via VFD. There isn't much posted about flowrates, but cutting head by half would only need a 25% speed reduction, so perhaps the required flow would still be within the VFD range too. I don't think the VFD or a mismatch of pump to system are causing the problems described.

The NPSHa being reduced at the higher flows associated with higher RPMs would be a symptom of poorly configured or missized suction piping, so I can agree with you there. A 25% increase in flow could certainly reduce NPSHa to unacceptable levels and we do know now that the suction piping configuration of one system is poor. The same problem is also occurring with different pumps-systems, something that might also suggest poor pipe design might be endemic at that installation. I've typically found that if poor piping configuration is discovered in one location, its likely to be found in other systems too, so I'd leave the VFD out of this one and go with NPSHa trouble.

http://virtualpipeline.spaces.msn.com

 

[drdoom]

Thank you all for your help. As I read your suggestions i see that there are many underlying concerns that I had never considered, however being the Yankee that I am I would tend to lean toward the simple yet maybe not the most obvious, the inlet piping configuration. As in most cases space was a premium and i was lucky to get enough room to be able to comfortably do O&M so not much was provided for piping, thus the inlets to both sets of pumps are configured as headers with "T"s or a 90 from the header direct into the pump eye. and as BigInch indicates i have two different pumping systems with totally different requirements, yet similar piping that are exhibiting the same problem.
i have a couple of questions:
1) Please define BEP, NPSHa and NPSHr.
2) Where might I find these curves? Is there a web site?
3) I am still wondering about the fact that when the pump is running at 100% and there is some form of cavitation occurring, the amp draw exceeds the nameplate data. as i stated before wouldn't this situation be similar to running the pump with no load and thus very little amp draw?

 

[hyposmurf]

1) Please define BEP, NPSHa and NPSHr.
BEP best deficiency point, seen at pump perfomance curve
NPSHa net positive suction head aviable , as per your instalation
NPSHr net positive suction head required , seen at pump curve

2) Where might I find these curves? Is there a web site?
your´s vendor web page
3) I am still wondering about the fact that when the pump is running at 100% and there is some form of cavitation occurring, the amp draw exceeds the nameplate data.

The cavitation occurs when the NPSHr is less than the NPSHa.


As i stated before wouldn't this situation be similar to running the pump with no load and thus very little amp draw?

As I stated before , at no load , but with water running , will demand more amp draw than at full load , that is an intrinsically feature of centrifugal devices as pump or fans.

Could you give us a mean to see yours pump brand and model , and best of all some dwg or photo of the installation ?.
you can upload it to

http://www.yousendit.com

, as we can not show here our e-mail , send the image and dwg or other data to yourself , the you will get a link , then put the link here , so we can see it .

 

[SeanB]

"In the case of the raw water pumps (maximum output = 900 GPM @91 FT of head)the rated amp draw is 44 amp and the 100% demand is 46+. Also i want to mention in the case of these pumps i have to limit the upper limit to 90% of speed or the pump goes into very noticable cavitation."

Normally when you specify a new pump, you define a "rated flow" - which should be your maximum flow. Your pump vendor then picks a curve that allows you to meet that rated point of flow and head. It sounds to me like you are trying to run your pump at "end of curve conditions" - in other words getting the absolute maximum flow out of the pump - which is not the right thing to do. Unless you stated that you wanted your pump motor sized for end of curve condition - it probably wasn't and hence the reason why you are running into the overamp situation. Also, if you are moving beyond the BEP on your curve, your NPSHr can increase substantially - this is probably where the cavitation is coming from. In my opinion you are asking for too much flow from the pump. Keep it at it's rated point and see what happens.