Need help with pump curve

[brianhamel]

I have a pump recirculating liquid around a scrubber. I want to know the flow rate at the current conditions (pump discharge pressure of 50 psig)

I know my head losses due to friction (10 feet) and elevation (40 feet) and I know the height of liquid above the suction side of the pump (14 feet).

Assuming the density of my liquid is that of water, how can I correlate the pump pressure with the head on the pump curve?

 

[katmar]

The pump discharge pressure of 50 PSI is equivalent to about 115 feet of water. Subtract the 14 ft that is due to the suction head and you can see that your pump is generating 101 feet of head. If you find this point on your pump curve you can read off your flowrate.

Strictly, you should decrease the 14 ft on the suction side by the friction loss there, but hopefully your suction piping is well designed and has negligible friction.

Theoretically you should also take the velocity head on the delivery side into account (i.e. add it to the 50 PSI that is "seen" by the gauge), but with liquids this can mostly be ignored.

 

[brianhamel]

katmar,

Thanks for the info. I'm thinking now that the pump discharge pressure is not going to be an adequate means of monitoring flow rate because of where I am on the pump curve. The pump curve shows capacities of 250 gpm at 100 feet of total head and 25 gpm at 110 feet of total head. These only translate to about 4 psi on my pressure gauge. Any suggestions other than installing a flow meter?

 

[sailoday28]

nomenclature
subscript s source of water supply
d discharge receiver
1 pump inlet flange
2 pump outlet flange


Z elevation
V velocity
p pressure (static)
rho density
(p/rho +Z +V^2/2)s = (p/rho +Z +V^2/2)1 +suct losses
(p/rho +Z +V^2/2)2 = (p/rho +Z +V^2/2)d +disch losses

Add the above two equations and rearrange
TDH =(p/rho +Z +V^2/2)d-(p/rho +Z +V^2/2)s + system losses

Where TDH is total dynamic head of pump.
Knowing supply and source pressure (atmospheric???) and suction and disch head (14 ft and 40 ft ????)
and their corresponding velocities (negligible ???)
Compute TDH. Neglecting density changes, use pump vendors H-Q curve H--- corresponds to TDH
and read Q.

 

[brianhamel]

Sailoday28,
Ok, let me get this straight:
My discharge pressure reading is 50 psig or 115 feet head.
My discharge elevation is 40 feet
My suction head is 14 feet
My friction losses are 9 feet
My velocity losses are negligible
TDH = (115 + 40 + 0) - (0 + 14 + 0) + 9 = 150 feet
Which, according to the pump curve is off the chart which - I assume means the pump can't deliver that much head. What am I doing wrong here?

 

[sailoday28]

brianhamel (Chemical)
1)If your suction source is from a large tank (therefore neglect its velocity head) and the liquid level, Zs =14 ft
2) If your final discharge is to a large tank (neglecting velocity head) Zd= 40 ft
3) And say the suction and discharge tanks have atmos pressure ps-pd =0
TDH= 40-14 +losses = 26+10= 36 ft.

For your pump disch pressure to read 50 psi, then something has to be wrong with the input 3) above. Do you know pd and ps?

 

[brianhamel]

sailoday28,

Ok, now we're getting somewhere. I'll scrutinize the system better tomorrow. Perhaps one thing I failed to mention was that there are spray nozzles on the discharge piping which will be giving me some back pressure. There must be more head losses than I have assumed.

Thanks and I'll be in touch.

 

[katmar]

Brian, you are correct that when your operating point is in the flat part of the curve it is hard to use it to get an accurate fix on the flowrate.

You might have better luck if you can get the flow characteristics of the flow nozzles. You know the discharge pressure at the pump. To get the pressure at the nozzles you will have to subtract the static head and the friction losses. Although you stated earlier that you know your friction losses, it can only be an estimate because you do not know the flowrate.

Using this estimate, calculate the pressure at the nozzles and from the nozzle characteristic calculate the flowrate. Compare this flowrate with your previous estimate and make a new estimate. Re-calculate the friction loss and do the whole thing over again. Repeat until the estimated flowrate does not change.

regards
Katmar

 

[rzrbk]

brianhamel,

In your equation above, the 50# discharge pressure is a measured value of what the pump is seeing, and will include any friction, elevation change, or backpressure in the discharge line. The 40' of elevation does not need to be added again.

The most accurate way to determine an existing pump's deltaP will be to use a single calibrated gage to measure both suction and discharge pressure directly. Then you don't have to worry about estimating piping losses, etc.

If the pump curve is too flat to estimate a flow from the deltaP, and a flow meter is not an option, then Amp readings on the motor leads will allow an estimation of power usage to help narrow the flow range down.