5 replies to this topic

[marthin_was]

Hi All,
I have several centrifugal pumps running parallely, and each pump installed with PCV to control pump pressure (pressure sensing located at upstream of PCV). We need certain pressure, but the flow requirement by users is variably will change. Is it necessary to install another PCV (pressure sensing at downstream of PCV) at common main header line? FYI, each pump discharge size is 14 in and main header is 30"

thanks,
marthin

[Padmakar Katre]

Hello
Dear Marthin
Please can you be specific
What is your concern to control pump discharge pressure
or the Main header pressure so that I can give a rough P&ID of the same
Now roughly I can tell you that to control pump discharge pressure either you have to put control valves in minimum circulation of every pump if you have to control the main header pressure then the logic will be somewhat different because as soon the consumption from the main header gets suddenly stopped then all your pumps will run at shutt-off head because there is no consumption now in this case all PCVs at pump discharge will start opening but that to a dead head as per logic cause your pump pressure is increasing.So if you want to control pump discharge pressure then you have to have PCV in minimum circulation line.The same P&ID I will send you

[marthin_was]

Hi, thanks for replying,
These are modification of existing pumps (please see attachment that show snapshot of PFD) and it doesnt have minimum flow return dedicated for each pump, but it has minimum flow return line for common header line. I use pressure information to open/close the minimum flow valve. The water need to be delivered with certain pressure at each users, and the flow requirement from each users are changes, so we have to keep that pressure setting stable. But in other hand, we want to keep the pumps running on it curve, so we will install PCV (valve A) at each disharge of pumps. But my contractor propose to install another PCV (valve at common header line. In my opinion its overkilling, because we already habe PCV A to control the pressure plus we have also valve C as safety device if minimum flow case happened. All sugestions from this forum are welcome,

thanks,
marthin

Attached Files

•  PFD__snapshot.xls   40KB   154 downloads

[djack77494]

marthin,

but it has minimum flow return line for common header line. I use pressure information to open/close the minimum flow valve.

I find several of the concepts shown on your flow diagram to be objectionable. You do not have any minimum flow lines at all, in my opinion. The header pressure control loop cannot really provide that function because any of the individual pumps may be choked by their individual valve "A" closing. So loop "C" only controls the discharge header pressure, which is not strongly linked to flowrates thru your pumps. My preference would be to measure the total flowrate coming from each of the pumps and to use this to modulate individual recycle control valves for each of the pumps. Then use a common user pressure control valve like your "B" loop to keep pressure steady.
Doug

[marthin_was]

Yup, I aggree and understand with your thought
But do I have another option besides installing minimum flow for each pump..?

[JoeWong]

marthin_was,

I have some comments on the following issue.

1) The PFD lookslike a Firewater network. Is it ?

2) If so, 2 SAFETY issues you may need to address.
a) There is single inlet to the system. Double inlets may required per code.
the PCV "B" may failed close due to reverse signal and leading to no flow to network. This is not acceptable from safety point of view.

3) Liquid is incompressible. With present scheme, i feel that PCV "B" abd "C" will swing with slight pressure change usptr. of PCV "B". Unstable supply is anticipated in the network.

4) The PCV "B" looks to me redundant as PCV "C" is provided to maintain the pressure.

5) Failed close on PCV "A" may lead to pump operate at zero flow and potential damage of pump(s).

6) Please check your pump curve. Present scheme and rather flat pump curve do not guaranteed minimum flow through each pump and lead to potential damage of pumps. In addition, pump discharge piping configuration may not be symetry. Eventhough your pump curve showed that your pump flow are well higher than minimum flow, mal-distribution may lead to high flow on one pump and low flow on the other pump. How the pump protection can be guaranteed with present scheme ?

My presonal preference quite inline with djack77494 proposal where to measure the total flowrate coming from each of the pumps and to use this to modulate individual recycle control valves for each of the pumps. Then use a common user pressure control valve like your "C" loop (instead of "B" loop to keep pressure steady. A buffer drum may need cater for instanteneous high demand.

Regards,

JoeWong