PD pump and control valve a redundancy? 1

[aspearin1]

I'm designing a test stand for high viscosity fluids requiring a positive displacement pump. I have drawn into my process flow diagram flow control valves for the system, but I'm having second thoughts.
In a normal operating low viscosity system, I would use a cetrifugal pump and control flow via a control valve. This makes reasonable sense to me since it is a constant pressure flow system. With a PD pump, it's a constant flow system, and most, if not all, PD pumps have variable flow control.
Does it make sense to put a control valve into a PD pump system? Will a partially closed valve reduce flow or just create an orifice through which the entire volume is forced, with an increase in pressure?

aspearin1

 

[jmw]

As you increase the back pressure by partially closing the control valve the pump pressure will increase thus maintaining a near constant flow.

This assumes that there is only the one flow path. If, however, the flow divides and there is a control valve in one flow path, then as the control valve is adjusted the flow balance will be modified.

If there is only one flowpath then if the valve closes completely disaster looms. Something will fail catastrophically.

Usually, PD pumps can be specified with an internal or external pressure relief valve as protection against this, it is set such that above a certain pressure the flow is recirculated back to the pump inlet.

You can see a rather different approach to this if you do a search for Plenty Pumps. They have a vane type design where the pump rotor cearance can be adjusted effectively allowing more or less slip between the vanes and the pump chamber and by this means adjusting the delivery rate.

The question is, what do you mean when you say: "...most, if not all, PD pumps have variable flow control." If you specify it, you can have pumps built with variabe speed gearing or you can specify a VFD to vary the motor speed.

 

[aspearin1]

The majority of PD pumps I've come across plug the use of variable motor speed for flow control. Thank you for your input regarding the addition of a second flow path. This is another approach I've been toying with and it appears to benefit the system in more ways than one. I will use a 3-way control valve to recirculate any "unused" fluid back to the tank. This should preserve the pump and possibly aid in mixing and temperature control.

aspearin1

 

[jmw]

Well now there is another thing; with high viscosity fluids even small temperature changes are going to lead to viscosity changes and this can mess up the flow balance aproach.
If the rate of flow through one path reduces, the temperature of the fluid will reduce because of increase heat loss and this will lead to a higher viscosity which will reduce the flow still further ......

 

[aspearin1]

The plan includes double-pipe heat exchangers on all flow paths (possibly to include static mixers) as well as a coil in the master mixing tank. PID control to a heater/chiller linked to an in-line viscometer. What I am concerned about is monitoring density changes. I have yet to find a reliable non-nuclear in-line SG meter. Though the rads on the meters are almost negligible, it would not be well received at my company, and would just end up a nuisance with our RSO.


aspearin1

 

[jmw]

You can get density and viscosity from the same device.
Two options are the Solartron Mobrey 7827 which is a tuning fork viscometer (0.5 to 20,000cP) and which provides a very good density measurement hence you get dynamic and kinematic viscosity plus the API base density and base viscosity (fixed and floating reference temperatures.
The second option is the new Coriolis based mass flow meter from Endress + Hauser which provides viscosity and, since it is a mass meter, also give density so you may get dynamic and kinematic from this also. However, as i am still awaiting information on this i can't do more than say it exists.
You need to visit the E+H web site and follow the mass flow links. Also visit

http://www.solartronmobrey.com

and

http://www.viscoanalyser.com