Bypass flow control 1

[andrewfreeman]

Hi Folks,

In our R&D test facilities we require water at very low flowrates (10 - 300 kg/h) and high pressure (50barg) for a series of experiments. We plan to use a CAT piston pump with VSD. However we still cannot achieve the lowest flowates (10kg/h). Therefore to achieve the lowest flows I plan to fit a bypass system on the pump outlet. I know this will be a waste of energy but it's research work only and won't be used that often.

My question is where does the throttling valve go. On the main line or on the bypass line? My thoughts are that it should go on the main outlet line, as that is where you want to control the flow. But my colleague thinks it should control the bypass line.

*Note: there will be an oval gear flowmeter on the main outlet to measure the flow.

Thanks in advance

Andrew

 

[Latexman]

We put a CV in the by-pass line to control pressure and a CV in the main line to control flow. Each loop affects the other so tuning will be critical. Pressure surges will be present if you do nothing about them. Stand pipes or surge dampeners on pump inlet and outlet will lessen them.

Good luck,
Latexman

 

[andrewfreeman]

Thanks for the reply Latexman

I think the water pressure will be controlled by the pump, so I will only require flow control. The water is being injected into a steam line (at pressure). The positive displacement will just pump against the resistance from the steam line. So I was led to believe the pressure will just 'sort itself out'.

Does putting in a bypass system mess up this self adjusting pressure control from the piston pump?

 

[iainuts]

I'd agree it goes on the bypass line. Being a positive displacement pump, if you put it on the main line downstream of the pump, you will simply overpressurize that portion of the line. Flow rate won't decrease simply by increasing backpressure.

If you put the valve on the bypass line, as you say "the pressure will just sort itself out". Control of the bypass valve still needs to be integrated with the VSD though. Valve control might best be done only once the VSD hits the lower limit of its speed.

 

[andrewfreeman]

bypass line it is then, my colleague will be happy he's right.

It's only an R&D lash-up so we will control the bypass line by hand.

Thanks

Andrew

 

[BigInch]

Bypass. Two valves on a recip is asking for trouble.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[dcasto]

You cannot flow control a PD pump by definition it is a flow control device. So, the flow control goes on the bypass. the only valves on a PD pump should be a maintenance valve on suction and discharge if needed a check valve, and a PSV in parrallel with the aformention flow control valve.

 

[hydromech]

The position of the valve really depends on the difference between the minimum flow rate required and minimum flow rate of the pump.

If the difference is small there would be no problem in putting a flow control valve in the pump line. A small pressure drop could be lived with..?

If there is a big difference in the flow rates, I would put a big valve in the by-pass line. Start with it fully open and then close it until you get the flow required through the process line.

This is all assuming that the usual safety features are installed to prevent excessive pressures.

 

[andrewfreeman]

The minimum flow rate of the pump is 43kg/h. We want to achieve 10kg/h.

The pump has a built in safety valve, also the pump outlet feeds water to a heat exchanger which has another safety valve on the hot water out incase of plate rupture.

Thanks
Andrew

 

[Latexman]

BigInch,

Most of our applications are multiple destination feed points. We've used a CV in the by-pass line to control pressure and a CV in the main line (at each destination) to control flow for many, many years. If we have a system with a single destination, we set it up the same way. Same control code and everything, because one of these days someone will want to tap into the header and use it somewhere else! And, we always put a spare nozzle in the header and a flanged off line run tee at the end of the line.

Two valves on a recip is not asking for trouble if you know what you are doing, but I agree that it is, if you don't.

Good luck,
Latexman

 

[rzrbk]

For what you are trying to do, a pressure control valve on the bypass line would probably be best. The flow meter would go in the main line downstream of the bypass branch with a globe or needle valve downstream to control the flow to the test.

As stated above, a PD pump will flow what it flows. The needle valve will control flow to your test, and the control valve will ensure the pressure you need and automatically dump the required additional flow.

Note: the bypass should not go directly back to the pump suction or it will start heating up the fluid. It should go back to the suction vessel, or be dumped to drain depending on the set-up.

 

[BigInch]

The pump will not flow what it flows, unless the downstream control valve and the bypass let it do so without overloading the rod.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com