13 replies to this topic

[Juan22]

Dear All,

 

It just came to me a design for a metering pump distributing additives to two reactors at the same time.

The design has two flow control valves on the two branches downstream of the pump, and one pressure regulator bypassing to the feed tank (please see attached). This design seems to me like a conceptual error because you usually don’t put a control valve on the main line downstream a PD pump. It doesn’t regulate the flow. Just increases the pressure losses. A waste of energy. Also, it seems to me that a pulsation dampener is missing at the discharge.

 

I would suggest a control logic regulating the flow of the main line (reactor #1) with the flow control valve in the bypass to the tank, and would put a flow control valve on the branch to the reactor #2. Also, I would add the pulsation dampener at the discharge of the pump.

 

I’m not an expert on PD pumps and would like to know your opinion about those two designs.

 

Are both right? Which is more suitable for the application?

 

Thanks in advance

Attached Files

•  PD-Pump-Two-Flow-Streams-found.pdf   41.45KB   68 downloads
•  PD-Pump-Two-Flow-Streams- proposed.pdf   30.02KB   54 downloads

[latexman]

"Found" will work.  We've used it for years in multiple plants.

 

"Proposed" will not work.  Throw that design away.

[Bobby Strain]

I think the proper design would use two metering pumps with variable volume, one for each reactor. And dispense with all the controls.

 

Bobby

Edited by Bobby Strain, 22 December 2017 - 11:52 PM.

[Juan22]

Thank you very much for your feedback Latexman and Bobby.

 

Regarding with the proposed one, may I know why won't work?

 

Bobby, I agree with your suggestion. Unfortunately, it's not my design.

 

Thanks,

Juan22

[fallah]

juan22,

 

The proposed one can work; provided that the demand of reactor #1 would be variable and the CV on the by pass line which is normally closed, would be opened to compensate the flow reduction to reactor#1. In normal conditions, the remaining of pump total flow after subtracting demand flow of reactor #2 will go to reactor #1.

[breizh]

Hi ,

Let you check with or for  Pulsafeeder with associated mass flowmeters !

I was using this brand when I was producing Latex .....

Accurate metering was  a must for this kind of process.

Agree with latexman and Bobby .

Hope this is helping you.

 

Breizh

Edited by breizh, 23 December 2017 - 03:26 AM.

[farid.k]

Juan22:

 

Agreed with latexmen.

Also practised in my plant. Each injection point must have the control mechanism (ft + control valve). The other valve that return back to tank (could be control valve+pressure transmitter or back pressure regulator) normally set slightly high than operating pressure incase say reactor one no need injection, pressure will build up and return to tank will open accordingly. And my plant has another relief valve that set higher for ultimate protection of the system.

You also may use 2 pumps (for dedicated injection point) or one pump with special injection device that call as injection rate control device (IRCD). It is actually same as per control valve with ft but they are slight smarter than that. You may check brand HASKEL, AMFLOW, SKOFLO for typical IRCD

[Juan22]

Fallah,

 

Thank you very much for your feedback. I agree

[Juan22]

breizh:

 

 

Thank you very much for your feedback and suggestions. I appreciate it. I'l check pulsafeeder.

 

I agree with you. That's why I'm surprised not having a pulsation dampener on the "found" configuration for better flow measurement.

[Juan22]

Juan22:

 

Agreed with latexmen.

Also practised in my plant. Each injection point must have the control mechanism (ft + control valve). The other valve that return back to tank (could be control valve+pressure transmitter or back pressure regulator) normally set slightly high than operating pressure incase say reactor one no need injection, pressure will build up and return to tank will open accordingly. And my plant has another relief valve that set higher for ultimate protection of the system.

You also may use 2 pumps (for dedicated injection point) or one pump with special injection device that call as injection rate control device (IRCD). It is actually same as per control valve with ft but they are slight smarter than that. You may check brand HASKEL, AMFLOW, SKOFLO for typical IRCD

Farid.K,

 

Thank you very much for the feedback and the suggestions. I appreciate it. I'm already checking on IRCDs!

 

I wouldn't imagine that a configuration going against the typical best practices for PD pumps is so popular (the "found" one). I know it works but at the expense of a waste of energy for the pumps.

 

I'm curious about the sizing criteria for the flow control valves. Typically, for centrifugal pumps the criteria is 10 psi or between 20-50% of the total pressure loss of the system.But I never heard about any criteria for PD pumps. Should we use the same criteria?

 

Thanks

[farid.k]

Juan22:

Agreed with latexmen.

Also practised in my plant. Each injection point must have the control mechanism (ft + control valve). The other valve that return back to tank (could be control valve+pressure transmitter or back pressure regulator) normally set slightly high than operating pressure incase say reactor one no need injection, pressure will build up and return to tank will open accordingly. And my plant has another relief valve that set higher for ultimate protection of the system.

You also may use 2 pumps (for dedicated injection point) or one pump with special injection device that call as injection rate control device (IRCD). It is actually same as per control valve with ft but they are slight smarter than that. You may check brand HASKEL, AMFLOW, SKOFLO for typical IRCD

Farid.K,

Thank you very much for the feedback and the suggestions. I appreciate it. I'm already checking on IRCDs!

I wouldn't imagine that a configuration going against the typical best practices for PD pumps is so popular (the "found" one). I know it works but at the expense of a waste of energy for the pumps.

I'm curious about the sizing criteria for the flow control valves. Typically, for centrifugal pumps the criteria is 10 psi or between 20-50% of the total pressure loss of the system.But I never heard about any criteria for PD pumps. Should we use the same criteria?

Thanks

For IRCD system, they have specific loss and my previous experience, for sizing purpose, loss across ircd itself is 7 barg for normal and 10 barg for worst case. Yea it sounds so big. for detail you may contact vendor for specific model sizing n pressure loss. The best about this things is you minimize the number of pump. This thing also can control at dcs, local and has manual by pass control.

[farid.k]

Juan22:

Agreed with latexmen.

Also practised in my plant. Each injection point must have the control mechanism (ft + control valve). The other valve that return back to tank (could be control valve+pressure transmitter or back pressure regulator) normally set slightly high than operating pressure incase say reactor one no need injection, pressure will build up and return to tank will open accordingly. And my plant has another relief valve that set higher for ultimate protection of the system.

You also may use 2 pumps (for dedicated injection point) or one pump with special injection device that call as injection rate control device (IRCD). It is actually same as per control valve with ft but they are slight smarter than that. You may check brand HASKEL, AMFLOW, SKOFLO for typical IRCD

Farid.K,

Thank you very much for the feedback and the suggestions. I appreciate it. I'm already checking on IRCDs!

I wouldn't imagine that a configuration going against the typical best practices for PD pumps is so popular (the "found" one). I know it works but at the expense of a waste of energy for the pumps.

I'm curious about the sizing criteria for the flow control valves. Typically, for centrifugal pumps the criteria is 10 psi or between 20-50% of the total pressure loss of the system.But I never heard about any criteria for PD pumps. Should we use the same criteria?

Thanks

Dear juan,

For flow control valve, That one you need to size based on according to typical control valve sizing. I am not so sure about the rules that you mentioned. be careful applying the rules as typically pressure drop across the system is minimum as flowrate is very small

Edited by farid.k, 23 December 2017 - 05:26 PM.

[Juan22]

 

 

Juan22:

Agreed with latexmen.

Also practised in my plant. Each injection point must have the control mechanism (ft + control valve). The other valve that return back to tank (could be control valve+pressure transmitter or back pressure regulator) normally set slightly high than operating pressure incase say reactor one no need injection, pressure will build up and return to tank will open accordingly. And my plant has another relief valve that set higher for ultimate protection of the system.

You also may use 2 pumps (for dedicated injection point) or one pump with special injection device that call as injection rate control device (IRCD). It is actually same as per control valve with ft but they are slight smarter than that. You may check brand HASKEL, AMFLOW, SKOFLO for typical IRCD

Farid.K,

Thank you very much for the feedback and the suggestions. I appreciate it. I'm already checking on IRCDs!

I wouldn't imagine that a configuration going against the typical best practices for PD pumps is so popular (the "found" one). I know it works but at the expense of a waste of energy for the pumps.

I'm curious about the sizing criteria for the flow control valves. Typically, for centrifugal pumps the criteria is 10 psi or between 20-50% of the total pressure loss of the system.But I never heard about any criteria for PD pumps. Should we use the same criteria?

Thanks

For IRCD system, they have specific loss and my previous experience, for sizing purpose, loss across ircd itself is 7 barg for normal and 10 barg for worst case. Yea it sounds so big. for detail you may contact vendor for specific model sizing n pressure loss. The best about this things is you minimize the number of pump. This thing also can control at dcs, local and has manual by pass control.

 

Thank you very much Farid.K.

 

It sounds Ok because the injection pressure is about 160 barg.

[Juan22]

 

 

Juan22:

Agreed with latexmen.

Also practised in my plant. Each injection point must have the control mechanism (ft + control valve). The other valve that return back to tank (could be control valve+pressure transmitter or back pressure regulator) normally set slightly high than operating pressure incase say reactor one no need injection, pressure will build up and return to tank will open accordingly. And my plant has another relief valve that set higher for ultimate protection of the system.

You also may use 2 pumps (for dedicated injection point) or one pump with special injection device that call as injection rate control device (IRCD). It is actually same as per control valve with ft but they are slight smarter than that. You may check brand HASKEL, AMFLOW, SKOFLO for typical IRCD

Farid.K,

Thank you very much for the feedback and the suggestions. I appreciate it. I'm already checking on IRCDs!

I wouldn't imagine that a configuration going against the typical best practices for PD pumps is so popular (the "found" one). I know it works but at the expense of a waste of energy for the pumps.

I'm curious about the sizing criteria for the flow control valves. Typically, for centrifugal pumps the criteria is 10 psi or between 20-50% of the total pressure loss of the system.But I never heard about any criteria for PD pumps. Should we use the same criteria?

Thanks

Dear juan,

For flow control valve, That one you need to size based on according to typical control valve sizing. I am not so sure about the rules that you mentioned. be careful applying the rules as typically pressure drop across the system is minimum as flowrate is very small

 

Dear Farid,

I agree. The pressure drop through the valves should be minimum.

Thank you very much