Pump Flow VS Pump Motor Amp 1

[Brian2903]

I am wondering if there are any relationships to descript pump motor Amp and actual pump flow. That is by measuring the pump motor Amp to eatimate the actual pump flow.

Thanks!!

 

[JRLAKE]

Yes. Given two points (power with impeller diameter) you can estimate the flow using the pump curve at the corresponding RPM.

Find where BHP curve and impeller curve meet, come straight down and you have your flow. You can double check the estimate using the differential head, if you have a way of measuring it. You also need to consider efficiency of the motor and other drive losses as well when calculating the power.

The accuracy of this approach is dependent on how many variables you include that affect pump curve accuracy, ie pump wear, impeller clearance, etc.

If accuracy is key, consider a flowmeter.

 

[TenPenny]

The short answer is, yes, there is.

What kind of pump are you talking about? Piston? Centrifugal? Gear?

 

[JJPellin]

The previous replies are absolutely correct. I would offer one word of caution based on an experience I had. Some big vertical turbine pumps actually have a hump in the Brake Horsepower curve. At extremely high flow, the amp draw actually drops at higher flows. This leads to a situation where one amp draw could relate to two different flow rates; one on each side of the hump. But this is very rare.

Johnny Pellin

 

[electricpete]

Pump has a bhp vs flow curve a was mentioned.

In very general terms:
-pure radial flow pumps have monotonically increasing BHP vs flow accross their range
- pure axial flow pumps have decreasing BHP vs flow accross their range
- mixed flow have a hump in the range

You also would want to consider motor current vs bhp. The simple approach accurate near full load is to use linear approximation current vs bhp load. But actual current is higher than this prediction when operating at low load. At 0 motor load, there might still be ~ 20% FLA current for 2-pole motor and 30% for 6-pole motor.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I should mention the linear approximation above would be based on nameplate FLA current and load.

If you have an accurate measurement of motor speed, and knowledge of line frequency you can also estimate motor load based on slip.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[BigInch]

Pssst. Clamp-on ultrasonic flow meter.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[waross]

Well pump horsepower may not be linear with pump flow.
Motor current is not linear with horsepower.
Changes in pump flow may change the dynamic head that will change the relationship between flow and horsepower.
You have to ignore;
1> The pump curve.
2> Dynamic head.
3> Motor power factor.
3> Motor fixed losses.
4> Motor load dependant losses.
Measuring the motor current will give an indication that the flow may be greater or less than it was, but it will take a lot of eaperience and history with the actual pump to get a reasonable estimate of flow, based on motor current.
Oh, and I forgot to mention fluctuations in supply voltage, and similar pumps supplied by feeders with different impedances may show different currents for similar flows. Admittedly a small difference.
I think that the answer is "Probably not, unless you have enough experience with this pump that you don't have to ask the question."

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

 

[BigInch]

Putting it in perspective, its like figuring how fast you're going from your gasoline consumption rate.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[JRLAKE]

I disagree with the previous two posts.

I estimate pump flow using amps and the pump curve frequently on centrifugal pumps. My accuracy is usually within 5%. Of course, there are exceptions, such as pumping high solids content, entrained air or non-newtonian fluids. However, for the most part, given the critical information, the relationship is quite predictable.

 

[BigInch]

And very steady state.


"If everything seems under control, you're just not moving fast enough."
- Mario Andretti
- when asked about transient hydraulic analysis

http://virtualpipeline.spaces.msn.com

 

[BigInch]

JR What do you mean you disagree with the previous 2 posts? I didn't say it was impossible. I know I'm going between 100 and 110 kph when my gasoline consumption is 9.5 L/h. I was just commenting on the fact that its a basackwards kinda' way to do it... IMO.


"If everything seems under control, you're just not moving fast enough."
- Mario Andretti
- when asked about transient hydraulic analysis

http://virtualpipeline.spaces.msn.com

 

[JRLAKE]

I mean, how do you think the manufacturers came up with the pump curves? Do you think they assigned a PhD M.E. to derive an equation using the geometries and CFD?

No. They hooked the thing up on a test stand, ran it and said, "Ok, someone write this down. At 3500 RPM, this pump is delivering X GPM at Y differential head and is drawing Z brake horsepower."

So, if you have Z HP, it makes perfect sense that you can work backwords, as long as you follow proper procedure working from the water reference point defined by the curve, consider reasons for possible curve inaccuracy (wear, etc.) and double check using a second point (differential head for example).

I swear we have a tendency to over analyze everything. But what fun would an engineering forum be if we didn't. I apologize for being cranky, laying people off is no fun and brings about a negative attitude.

 

[JRLAKE]

"JR What do you mean you disagree with the previous 2 posts? I didn't say it was impossible. I know I'm going between 100 and 110 kph when my gasoline consumption is 9.5 L/h. I was just commenting on the fact that its a basackwards kinda' way to do it... IMO."

I misunderstood your point was Big Inch. It is backwards I agree. That's why flowmeters were invented.

I don't think I could estimate my car's MPH by looking at fuel consumption, but at the same time, I ain't that smart either.

 

[JRLAKE]

...average fuel consumption I mean...different story most likely with instantaneous fuel consumption...I could do that...I see what you are saying now...which brings me back to a point I made in another thread...you have much cooler toys than I do...

 

[rmw]

I have found that if there are other means of determining pump flow such as flow meters (in who knows what kind of calibration) pressure gages, etc and a good cross check method for verification is needed, the pump amps anticipated by the pump curve at that point can either verify the data or bring it into question. I often rely on the amp readings although you have to have some reasonably accurate information on the PF of the power to the pump. As I normally used this technique in power stations, that information was usually at hand.

rmw

 

[waross]

I stand by this statement.Measuring the motor current will give an indication that the flow may be greater or less than it was, but it will take a lot of experience and history with the actual pump to get a reasonable estimate of flow, based on motor current.
Oh, and I forgot to mention fluctuations in supply voltage,
No offense taken JRLAKE. I've had bad days myself.
The point is if you are familiar with the pump, and you obviously are, you may get an indication but be wary of readings on strange pumps until you have some second source confirmation of your current based estimates. Actually a wattmeter gives a better indication. Some engineers do a shut-in load test and subtract that reading from the flow reading to factor out pump losses. Depends on the specific application though.

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

 

[Artisi]

I would say that to get somewhere near to a reliable flow using input current to a motor you would have consider and factor in ALL the variables effecting the motor output power and also establish very accurate differential head, this could only then be applied to an actual test curve for that particular pump, this is also dependent on the pump being in exactly the same configuration and internal condition as when originally tested and with ideal flow into the pump etc etc.
Anything else becomes a dream and wishful thinking.

 

[JRLAKE]

"That is by measuring the pump motor Amp to estimate the actual pump flow."

The question is can the flow be estimated, not calculated. I guess we all have different definitions of estimate.

I am also making the assumption, and probably incorrectly, that he is attempting a "one-time" quick estimate. If he is trying to implement a permanent system to continuously estimate flow, amps is definitely not the way to do it. But I assumed if he were trying to do this, he wouldn't even consider the amp approach and would be asking for recommendations of flowmeter vendors.

Since he is apparently looking for a non-intrusive and simple way to estimate flow, I assume he is looking for a quick glance at estimated flow. However, I will acknowledge that I frequently make the mistake of assuming others have taken a logical path to the obstacle that produces the question.

 

[rmw]

This is anecdotal and only on one set of pumps but I had a situation once where I had a significant variance between the pump flow indicated by the pump curve based on the differential gages and the outflow weir readings. They were significantly different than the outflow weir readings indicating that the pump flow wasn't enough (to maintain the measured condenser vacuum in a power plant turbine condenser.) The pump differential readings were taken over several operating points (river levels) and corrected to pump centerline. They were even taken pre and post pump(s) overhaul. The motor amps always verified the differential pressure readings and the pump curve.

As an aside, the piping was such that there was no location accessible to use a Doppler meter and other flow measuring devices were impractical or impossible (96" piping). It was known that the condenser outflow piping was separated and it was hypothesized (and strongly suspected) that a portion of the flow from the condenser found its way back to the river via the subterranean drain piping system put in place to drain the original plot when it was filled in from the river. There was good evidence for that. The fill dirt on which the plant site was located was obtained by building a drainage system and pumping muddy river water in and letting the water drain out after the sediment settled out. The plant major pieces of equipment were built on pilings. A civil engineering buddy of mine showed me photos of the original site with the drain piping in place. I believe that the flow from the broken main found its way to the original drain piping and by passed the weir. The motor amps proved to be a good verification of the pump differential and heat transfer calculated flow rates.

And, yes, artisi, I has to take every variable in the motor HP formulae into account and factor them in too. But in the end, it was 3 against 1 - pump curve, motor amps and heat transfer calculations that all agreed - and the weir lost.

rmw