MEASUREMENT & ESTIMATION OF PUMP FLOW

[freefallingbody]

Hi,

I have a 25 mWC, 510 m3/h rated pump directly coupled with 60 HP, 1470 rpm motor pumping water in a cooling tower circuit.

I wanted to measure the flow of this pump. I used a transit time type ultrasonic flow meter and got a flow of 520 m3/h. Since I don't trust these ultrasonic meters, I wanted to cross check the flow by another method. Since this is a new pump, I thought of cross checking the flow by measuring diffential pressure and refering to performance curves.

I measured the differential pressure across the pump. It was found to be 24 mWC. From the performance curve of the pump, corresponding to the head of 24 meters, the flow is around 525 m3/h. Both flow measurement and estimation are thus very close. I was very happy.

...But then I found out that the power input to the motor is only 32 kW. This would theoretically make the pump more than 100% efficient. I was expecting a motor power consumption of 43 to 45 kW based on the pump performance curves.

Any thoughts on what to do next, to decide on the flow?
(I even went to the extent that I switched off the pump from the main electical panel to ensure that I am measuring the power input to this pump only!)

Thanks in advance.

Dinesh

 

[tstead]

You measured flow by two different methods and arrived at essentially the same value. I would therefore suspect your power reading.

A couple thoughts, though...how did you compute input power? Did you take into account the motor's power factor or the effects of three-phase power?

Tim

 

[25362]

Following tstead's post:

Three-phase kW=0.00173(volts)(amperes)(power factor)
One-phase kW=0.001(volts)(amperes)(power factor)

 

[freefallingbody]

Hi,

I have measured power using a 3-phase power analyser ( Elcontrol, Italy make), which gives 3-phase power reading directly. I got 32.7 kW. I must have repeated this measurement 2/3 times.

From the independant voltage, current and power factor readings also I am getting the power of 32.9 kW. This deviation is acceptable.

Line Voltage = 405 Volt
Line Current = 55.8 Amp
Power input = 32.7 kW
Power factor = 0.84
Frequency = 49.7 Hz
(In India, we have line frequency of 50 Hz)

Thank you very much for your responses. Looking forward to hearing some more likely errors,

Dinesh

 

[hydrae]

Dinesh

So back to your original question, how else to determine flow.
I assume the flow is:
draw from cooling tower, pipe, pump, pipe, condenser, pipe, cooling tower

look at the piping, have you got any run long enough to a calculate pressure loss?
look at the condenser does the manufacturer have pressure drop curves for tube bundle- header system
look at the cooling tower are all the spray nozzles of a documented size and flow pattern?

A detailed analysis of any these components will give a pressure drop. Or combine all of the analysis for an overall pressure drop and compare to the pump curve.

Another method is to install an insert meter, available in both paddle wheel and magnetic.
Insert meters typically use a 2 inch tap (50mm) and can be done live then the meter is inserted into the flow. Both the paddle wheel and electro-magnetic produce a 4-20 mA signal and cost about $500 to $2000 usd.

Also how straight was the piping in the location you used the UT flow meter? you want 5 diameters or more upstream and at least 2 d downstream, the insert meters are also fitting sensitive.

Hydrae

 

[checman]

Siphon effect in the system ??? Regards Checman

 

[quark]

It seems Checman is close to the answer. What is the configuration of piping system?

 

[25362]

The location of the pressure gage may introduce errors in the estimation of the differential head developed by the pump: if measurements are carried out on larger diameter sections of a pipe, the velocity head drops o/a of the pressure head (law of conservation of energy) resulting in a higher pressure reading and viceversa.

Besides, are the gages calibrated, or, have you tried to use the same gages for suction and discharge pressures ?