Measuring Pump Header Flow Rates 3

[IBechir]

Greetings

For centrifugal circulator pumps operating in parallel, isn't important to be able to verify the flow for each pump individually?

I know that given the pressure gauge readings and the pump curves the flow can be determined but I'm asking about verification of flow of each pump while operating in parallel.

Is it enough to measure the differential pressures and the RPM or is it possible for a pump to produce an unexpected flow due to damage to the impeller?

This is an inquiry regarding general practice as I have never come across a design showing pump headers that have flow meters for each pump.

Maybe I am answering my own question but if had to verify flow I would use the flow meter on the main return pipe an take the reading when only one pump is operating.

Much obliged for sharing your expert opinions.

 

[LittleInch]

It kind of depends on the size of the units and how they operate, but yes, in parallel pumping, even with "identical" units and piping set up, you would be very lucky to get an exact split.

If your pump curve is fairly steep at the flow rate where the pumps work then it shouldn't be too bad - maybe 5 to 10% difference, but if you end up working at a point on the curve where it is much flatter it could be 50% or worse.

Usually the easiest method is to be able to compare running current to see which pump is talking the lions share of the work.

what each pump dies when it is working on its own is no indication of what happens when a second or third or fourth unit is operated in parallel. Of course all these pumps should have a common inlet and common discharge otherwise all bets are off.....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[KMenges]

LittleInch's response is good. Power monitoring works reasonably well to determine flow splits. I would add to that yes, you can measure individual flow from each pump from the return header while its operating on its own, however that is not the flow rate it will produce when it is operated in parallel. If you need to verify the individual flows for each pump while they are all operating I would recommend renting a portable ultrasonic flowmeter and measuring each pump discharge flow while they operate in parallel.(Assuming there is a steady state total flow at the header of course) If any of the pumps is producing less flow it may have worn internals or improper impeller clearance setting which essentially causes the pump head curve to be lower than design, which forces that individual pump to operate further back on its curve.

Not sure how common it is to measure individual flows in circ systems but it is very common in many industrial applications.

 

[IBechir]

Thank you. For general monitoring purposes of an HVAC closed-loop system measuring the running current sounds very reasonable. I read somewhere that the kW input is also an indication of the wire-to-water efficiency. I suppose is there is suspicion that one of the pumps may be operating poorly the flow can be verified with a portable ultrasonic flowmeter, like you said.

 

[EnergyProfessional]

You cannot measure flow by measuring pressure gain over the pump. At least not if you care about accuracy. The pump curve is very flat and the measurement very inaccurate. Look at a pump curve and see that just 1 psi difference can move your "measurement" from 40 to 70 gpm or so... (your pump will be different than my example). Now look a that gauge that the pump has built on and you see this is not accurate at all.

Alternative methods are balancing valves (caution, when measuring you are reducing the flow since you need a large pressure drop to measure) or some flow meter. If you have enough straight pipe you could use one of those ultrasonic strap-on meters - they are very accurate.

 

[IBechir]

The pumps I usually encounter have variable speed drives specified and no balancing valves on the individual pump discharge pipes; waste of energy. Finding enough straight distance for flowmeters, on the individual pump headers, could be problematic.

 

[EnergyProfessional]

if these are same size pumps, you could measure elsewhere and divide the flowrate.

unless you want to show the pumps don't work as advertised, why do you want to measure the flow? I've never seen a situation of too little flow where the pump was to blame.

 

[IBechir]

A magnetic flowmeter is usually specified on the main building return. So yes, its very possible to measure the common flow and the pumps are identical so I might assume that they are operating at equal capacities but for commissioning and perhaps for troubleshooting its a good idea to verify that the identical pumps are operating at equal capacities.

After reading the replies in this post I think measuring the current input, the shaft RPM and taking the readings from the pressure gauges at suction and discharge should be enough to verify that the pumps are operating at equal capacities- at least for HVAC chilled water applications.

A single DP sensor across the parallel pumps installation is more accurate than the pressure gauges but this reading will include the losses from common header to pumps (pipes + valves).

 

[LittleInch]

Sounds good to me.

I would program into the maintenance schedule a check on each unit running alone every 6 months or so to see if any unit is significantly less than any other and also to plot any changes over time.

Ditto a check on the running amps on a regular basis, once a month?

I would say a normal range of operating current in parallel mode is 10%. Anything more should be investigated

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[EnergyProfessional]

By measuring these you may be able to tell if they operate equally... but not how much flow in each exactly.

Again, not sure what you want to accomplish. If a pump fails it stops... there is no slow degredation unless you believe the stainless steel impeller wears out. And unless the manufacturer installed the wrong impeller, the pump will perform as advertised (unless you buy a chinese knockoff, like those Glundfos or Cinosset pumps

If you have a lack of flow it is a design error, or contractor installed too small pipes etc. It isn't the pump.