Control of Feed & Bleed Membrane System for Product Concentration

[burndirt]

thread136-153629

http://www.eng-tips.com/viewthread.cfm?qid=153629&page=2

Continuing from the thread above, does anyone have any advice on how to control a multi-stage membrane concentration system for continuous flow production?
I need the permeate + product flow to match the (moderately) variable inlet flow.

I need a high degree of concentration (remove ~80% of the volume) so I may need 2-3 stages or more. Liquid is water.

How did the control schemes mentioned in the old thread work-out? What are the best configurations for this that minimize overall capital cost and difficulty of control?

 

[FiltrationEngineer]

Can you use an equalization tank to take out the variable inlet flows and then run the system at a constant feed and bleed flow rate?

The control strategy woud include a constant feed flow and constant permeate rate, controlled with a control valve on the permeate and a vfd or control valve for the feed.

You'd need good PLC controls to prevent overconcentrating the feed side and fouling the membranes.

I'm not really clear about the multiple stages and what is required there.

You could also consider a strategy where the PLC figures out a volume of feed/permeate/bleed that happens in the system. For example, for every gallon that is introduced into the system, .8 gallons goes out as permeate at a controlled max flux rate while the .2 gallons goes out as concentrate.

Keep in mind that the lower the flux rate, the better the operation of the system.

 

[membrane20]

Feed and bleed system is good for concentrating 5 times or more, but this is less efficient than multi stage filtration at high flow rate.

If feed flow can be handled with less than 2-3 channels, feed and bleed system might be more economical. If you need more than 4 channels, you may consider a multi-stage system.

For better advice, I need solids characteristic, feed flow rate, approximate relation between flux and solids%.