5 replies to this topic

[rsk]

A reciprocating type positive displacement pump is to be used for applications where a precise control of pressure is required in discharge. Suppose Q be the flowrate of pump & Q1 & Q2 are the minimum & maximum flows required in the downstream applications. A recirculation line from pump discharge to suction with a control valve is proposed for pressure control at 120 kg/cm2. The pressure at pump suction is 4kg/cm2 now what are the broad steps for sizing of a control valve in this case.The line sizes from pump discharge to CV & form Cv to suction are 3m 2"sch80 & 2m 3"sch80. What is a criteria when i shall go for two Cvs in series.
regards

[djack77494]

rsk,
It's not clear whether you are interested in one or two control valves, but I can tell you that it doesn't matter. Neither will work. You do not want to put a control valve on the discharge of a PD pump, intending to control the pump's discharge pressure. A recip-type PD pump generates pulses that are shaped like half sine waves. You can think of those pulses as being flowrate or pressure, in doesn't matter. You'd be accurate in both cases. You do not want to use a control valve in the discharge to control either flow or pressure, since you cannot effectively do this in pulsating flow. Remember, half the time (for a simplex pump) the flowrate is zero. How do you control pressure then? You don't, unless you have an intermediate device, such as an accumulator, that can effectively smooth out the pulsations and produce a nearly steady flow (and pressure). While accumulators are somewhat effective, this is a tall order. I'd highly recommend that you consider alternate solutions or you relax your requirement for such steady pressure.

[Qalander (Chem)]

rsk,
It's not clear whether you are interested in one or two control valves, but I can tell you that it doesn't matter. Neither will work. You do not want to put a control valve on the discharge of a PD pump, intending to control the pump's discharge pressure. A recip-type PD pump generates pulses that are shaped like half sine waves. You can think of those pulses as being flowrate or pressure, in doesn't matter. You'd be accurate in both cases. You do not want to use a control valve in the discharge to control either flow or pressure, since you cannot effectively do this in pulsating flow. Remember, half the time (for a simplex pump) the flowrate is zero. How do you control pressure then? You don't, unless you have an intermediate device, such as an accumulator, that can effectively smooth out the pulsations and produce a nearly steady flow (and pressure). While accumulators are somewhat effective, this is a tall order. I'd highly recommend that you consider alternate solutions or you relax your requirement for such steady pressure.

Dear rsk,hello,
Although much is very correctly pointed out by djack, at times it helps if the service application& process is elaborated.
Moreover the very concept of reciprocating motion of pistons/plungers/rods whatsoever can not permit smoothness in flow and/or pressure downstream of pump in question;you should realize.
Best regards
Qalander

[JoeWong]

Doug has rightly pointed out pulsating flow on recip-pump and accumulator to smoothen flow. This kind of system needs proper engineering to ensure a good control...

Dropping from 120 kg/cm2(g?) to 4 kg/cm2(g?) is likely to occur and potential lead to valve cavitation when it pressure recover from ?? kg/cm2(g?) in vena contracta (VC) to 4 kg/cm2(g?). Anti-cavitation trim or multiple CV may solve the cavitation problem...

[djack77494]

I don't want to get stuck on the point, but I did want to clearly indicate that it is folly to try to use a control valve in pulsating flow service. Yes, you can use an accumulator to (somewhat) smooth out the flow and pressure, but you'll still maintain some of the pusating nature of your flow. A PD Pump itself is a flow control device (which is why they are sometimes called metering pumps). No additional flow control should be necessary. Many if not most PD pumps have adjustable strokes and/or speeds. If you want to control a process variable that depends on the flowrate through a PD pump, I would recommend that you control the pump's stroke or speed. I highly discourage any attempts to throttle the flow from the pump.

[JoeWong]

I don't want to get stuck on the point, but I did want to clearly indicate that it is folly to try to use a control valve in pulsating flow service. Yes, you can use an accumulator to (somewhat) smooth out the flow and pressure, but you'll still maintain some of the pusating nature of your flow. A PD Pump itself is a flow control device (which is why they are sometimes called metering pumps). No additional flow control should be necessary. Many if not most PD pumps have adjustable strokes and/or speeds. If you want to control a process variable that depends on the flowrate through a PD pump, I would recommend that you control the pump's stroke or speed. I highly discourage any attempts to throttle the flow from the pump.

Noted your concern.