Instrumentation Along Water Line for Flow and Head

[Wuest]

Hello,

I am building a system to test hydraulic conductivity in situ. A big component of this system is a water line that runs from the surface to some depth below the surface. The line will go from a water tank, to a pump, to a flow gauge, then to below the ground. My question is, would there be a way to determine the head added by the pump at any point in time (with constant flow and gravity head measurement) without fitting two pressure gauges on either side of it? I thought about using the pump performance curve to relate the rpm of the pump and the measured flow to the head given by the chart, however I believe that chart was developed assuming the water is being pumped from zero head to some head above that. In this case, the water would already be flowing even if the pump was off, so another question would be, does a performance curve show the amount of head a pump adds at that flow or does it show the amount of head added by the pump assuming the pump is the only factor causing the flow?

I would really like to get away with not purchasing two pressure gauges if I do not need to.

Any help or references you could point me towards would be greatly appreciated.

Thanks

 

[LittleInch]

So long as you know the flow rate and pump speed then with a good pump curve you can estimate the differential head / pressure imparted by the pump without gauges.

It's not ideal but should get to you to within 10%.

This is for a centrifugal type pump.

If yours is a positive displacement type pump then it won't work.

The fact that you can get flow without the pump is irrelevant. The pump doesn't care and if it is acting as a booster pump then so long as all the flow goes through the pump ( why wouldn't it?) then the pump will simply add head.

The pump curve is always a differential head. In reality for testing it almost certainly is fed with water at more than atmospheric pressure but they just subtract the inlet head from the outlet head to give you the differential which is what is shown on the pump curve.



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

 

[Wuest]

LittleInch,

Thank you very much for the response, that is great to know, I looked at a lot of pump curves but never found full definitions of the variables. It is a centrifugal pump so that method would work. However, now knowing that the accuracy will most likely be +-10% I will probably end up using pressure gauges anyway.


Thanks Again

 

[LittleInch]

The 10% is just because pump curves are rarely super accurate for your particular pump and other things like temperature, wear on the pump etc mean differential pressure is not the most reliable of numbers to pluck from a curve vs flow.

However if you've tested the actual pump you're going to use and the curve at the flowrate you're using is fairly flat then you might get <5%. bigger pumps vary less also so if this is quite small and low power the accuracy of the curve is also less.

gauges are best though.

LI

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

 

[ashtree]

Following on from LI if you have one pressure gauge on the output, a throttle valve and a flow meter then you could construct your own pump curve by plotting the relationship between head and flow.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"