Loading versus slip

[77JQX]

Hi - I hope you all can help me out. I'm a civil, trying to understand a centrifugal pump's operation - it isn't operating on the published curve. I'm wondering if I have an accurate curve, or if I have a motor problem. Mechanically, the pump is in good condition - we've inspected and measured the impeller. The motor, on the other hand, is a bit overloaded. The motor is a Reliance "Type P" 30 HP NEMA B with a 1.1 SF and nameplate rpm of 3530. The motor was installed in 1979. Based on the curve efficiencies, the motor is producing on the order of 30-32 HP whenever it is run. Typically it has one start per day, and operates continuously for 1 to 3 hours per day.

So here are my questions: For the pump curves to make sense, this motor is operating at about 3380 rpm (6.1% slip). Do you think this much slip is likely, or should I suspect my pump curves? Or is my issue a combination of motor speed and inaccurate curve? When I get my maintenance people back out there, what should they measure?

Thanks for your time.

 

[electricpete]

The motor can run lower at higher speed than “expected” due to
1 – overload. However if that much overloaded, you’d probably trip.
2 – low or unbalanced voltage. Slip is proportional to V^2 / Rrotor
3 – mis-connection causes the winding to see a low voltage even though system voltage is normal. For example connecting delta motor in wye.
4 – Rotor defect. Causes Rrotor increase. Again Slip is proportional to V^2 / Rrotor

fwiw, My gut feel is that you will find none of the above 3 applies to your situation.In similar scenarios at our plant, the problem usually lies on the side of understanding the fluid / pump characteristics, not the motor. My gut feel is none of the 3 applies


=====================================
(2B)+(2B)' ?

 

[electricpete]

The first sentence should have read:
The motor can run at higher slip than "expected" due to...


=====================================
(2B)+(2B)' ?

 

[petronila]

Hello 77

The slip increase with load and is necessary for torque production.

So I think the slip is normal if the motor is a bit overloaded like you sayd.

Regards

Carlos

 

[Compositepro]

How accurate are all your instruments? I can't say if 6.1% slip is excessive for your motor but it isn't far off from what to expect. I doubt you know your flow rate and pressure to better than 2-3% accuracy. It sounds to me that your measurements fall within the error band. Why don't you measure the motor speed? That would be far more accurate that trying to infer it from the pump curve.

 

[electricpete]

Why don't you measure the motor speed? That would be far more accurate that trying to infer it from the pump curve.

I agree. Also, if the cause is overload and your motor is powered from 60hz (not vfd), then your (op's) motor is operating at about 3 times full load. Check current using clamp-on and observing applicable electrical safety precautions.

=====================================
(2B)+(2B)' ?

 

[77JQX]

Thanks guys. I'll get the mechanics to check the speed - I just wanted to see if that was reasonable before I sent our already short staffed maintenance dept on a wild goose chase.

Electricpete - what do you mean operating at 3 times full load? Yes, we are operating at 60 Hz.

 

[electricpete]

what do you mean operating at 3 times full load?

EstimatedLoad = (SyncSpeed – ActualSpeed) / (SyncSpeed – NameplateSpeed)
EstimatedLoad = (3600 - 3380) / (3600 -3530) = 314%

This is a rough estimate, subject to some uncertainties errors mentioned above (voltage, slight rotor resistance variation with temperature, uncertainty in knowing true line frequency, possible measurement errors, possible errors in manufacturer labeling the nameplate etc).

I would not believe this until I had verified with currrent clamp-on that the motor is drawing approx 3 times full load amps.

If you have developed this estimate of slip by trying to tweak your pump curve to match your operating point using DP~N^2, Q~N.... I think the “error” likely lies on the fluid side of that calculation.


=====================================
(2B)+(2B)' ?

 

[77JQX]

Thanks electricpete. I have little confidence in the manufacturer's curve, but the deviations I had seen could also be explained by a speed difference. What I understand from your posts is (1)speed is be roughly linear with load in the overload region, and (2) if the motor & electric supply is in good condition, then a 6% slip is not credible. Thanks again.

 

[electricpete]

Yes, that’s a good summary.

=====================================
(2B)+(2B)' ?

 

[KllrWolf]

The quickest way I know of to estimate about where a motor is running load wise is to check the power into the motor and see how it compares to the rated power (taking into account power factor and efficiency). If both your voltage and current are close to rated (both within about 1% of full load rated) you are not overloading the motor.

Something to look at is are you using tested curves for both the motor and the pump, or just the typical curves? Most typical curves are generated purely by theoretical calculations. Some typical curves are created by making a perfect part and doing everything to make it operate the best, no matter the cost (then production is just slapped together).

Has the pump/motor been rebuilt since the curves were published?

One other thing, how long has the operation been off the curve?

 

[jraef]

The motor is a Reliance "Type P" 30 HP NEMA B with a 1.1 SF and nameplate rpm of 3530. The motor was installed in 1979. Based on the curve efficiencies, the motor is producing on the order of 30-32 HP whenever it is run.

So you have a 30HP motor with a 1.1 Service Factor, meaning it is technically capable of 33HP, and it is running the pump at 30-32HP. I don't see the problem here.

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[7anoter4]

My slip formula is more modest than of electricpete. From approximately:
T=2*Tbr/(slipmax/slip+slip/slipmax) I got :
slipmax=sqrt(Tbr/T*slip^2-slip^2)-Tbr/T*slip [Tbr=Breakdown torque]
Taken Tbr=2[*Trated] slipmax=7.26% and returning to the first equation T=1.97*Trated. That means 97% overload and it could not work in this condition.
You have to check the no-load slip and the actual current under load. I agree with jraef: if nothing unusual is happened with your motor -no overheating, no overload trip-then only a wrong measurement could be. Check also the hydraulic data: flow [GPM], head [FT] and specific gravity. BHP= (GPM*H*Sp.Gr.)/3960/eff.
If you take the pump velocity from v^2=Head/2/g and v=rpm*D[ft]/229 [ft/sec] D=diameter of pump rotor.
Then rpm=229*v/D. The error from all this [and mainly from sp.gr depending on temperature also!] could be more than 6%.Pump efficiency could be a source of error, too.

 

[electricpete]

Thanks 7anoter4 - your formula would be better for slip this high.

Sounds like we are all in violent agreement regarding the conclusion.

=====================================
(2B)+(2B)' ?

 

[7anoter4]

Thanks, electricpete. In fact this is what you already said:" I think the "error" likely lies on the fluid side of that calculation". I developed it only, a little bit more.

 

[ozmosis]

77JQX
how are you measuring power?
if your motor nameplate indicates 3530rpm at rated frequency and rated nameplate voltage then the current drawn 'should be' the rated amps on the nameplate if the motor is running at 3530rpm. If the motor is operating at rated frequency on rated voltage and running at 3380rpm, the current will be higher due to the motor running in overload. What amps is it drawing?
I assume the pump is DOL (Direct on Line) i.e. no VFD controlling or electronic softstart or fluid coupling.

 

[electricpete]

OP stated "Based on the curve efficiencies, the motor is producing on the order of 30-32 HP whenever it is run."
So the power was estimated from fluid parameters.

=====================================
(2B)+(2B)' ?

 

[waross]

With a rated speed of 3530 RPM you are looking at a normal slip of 1.94%
A slip of 6.1% is unrealistic. This equates to an overload of more than 300% by simple calculations. No need to get into motor torque curves for a more exact figure. At 6.1% slip, the motor would be so overloaded that it would quickly be shut down or destroyed.
Check your pump, dynamic head, specific gravity, impeller damage. Anything else you can think of.

Bill
--------------------
"Why not the best?"
Jimmy Carter