3 replies to this topic

[go-fish]

Hi

If centrifugal pumps (2 x 100%) in parallel are selected just to transfer contents from one tank to another, is there a requirement to have controlled min. pump recirculation line and a flow control valve in the pump discharge?

My understanding is that if the pump is selected for the normal flow (based on time to empty the source tank), it will just pump more at the start when the destination tank is empty and pump less towards the end when the destination tank is full. There is no need to control the flow.

Thanks

[Art Montemayor]

You are making an error because you are generalizing when you should be considering the specifics of the application. First draw a sketch of your application and you will see that you can make the transfer from the bottom of the source tank to either the top or to the bottom of the target tank. Which will it be?

Now look at your pumps’ performance curve. If you do not throttle the discharge of the pump, you will go to the “end of the curve” when pumping to the bottom of the target tank initially. This will probably over-draw more current than your drive motor can take and the motor will shut down on overload. If you pump to the top of the target tank, you may have sufficient static head to overcome that you may put enough head pressure on the pump – but that depends on other factors as well --- the distance and height between tanks, the size of the piping, the total frictional and system pressure drop, the performance curve characteristics, the size of the pump and motor starter, etc., etc.

You just can’t make a general statement and expect a specific answer. Engineering can’t work that way. In practice, there should always be a throttling or flow control on the pump’s discharge when transferring liquid with a pump.

[go-fish]

Art

Thanks for your response and apologies for not thinking like an engineer for a moment.

I just looked at the issue by drawing system curves. It may be possible to do it but will require a pump with a drooping performance curve. Due to this only a specific pump can be used which may offset the CAPEX saved by eliminating 8" control valve. Additionaly the pump efficiency will fall for the extreme static head difference cases.

I will go with the best practice of using flow control valve in pump discharge.

[S.AHMAD]

Go-fish
1.For throttling, you can use globe valve and estimate the flow from the pump curve and the pump suction/discharge pressure.
2. The general statement that pumping rate is more at the beginning and less at the end is very true. This is due to the fact that the source tank level is dropping (hence suction pump pressure reduces correspondingly). The destination tank, if the inlet at the bottom, the pump head required increases as the tank level increases. However, it the inlet is at the top of the tank, level in the destination has no effect (remains constant)
3. The above comments is a general one. As mentioned by Art, you need to be more specific and look/study the situation case-by-case.