How do you control a two pump cross flow filter system?

[mechprocess]

This thread is dedicated to a two-pump cross flow filter systems. All experienced comments are welcome regardless of system functional issues.

I have a system that uses a feed pump and a recirculation pump, a cross flow filter and back pulse tank. The slurry is feed into the filter loop system that is maintained at a pressure using a bleed back control valve. (Newtonian, 5 wt % 0.1 micron, ).


Does anyone know how to make one of these work?

 

[hprorelis]

Dear sir
recirculation pump is used to beat modules pressure drop and generate appropriate linear velocity.
Feed pump is used to control inlet module pressure
Best regards

 

[Jigneshpadia2002]

Two pump crossflow filtration system is uesd when one need to control the flux and reduce the fouling of the membrane. Typical application included cell harvest and protein purification.

Whenever the feed is fouling or requires the a low pressure constant flux application one needs to use a two pump TFF system. TFF = crossflow filtration.

Another appliation of two pump system can be a cascade operation. where the permeate of the previous unit can be feeding to the subsequent TFF system.

Feed pump to the membrane needs to be operated in such a way that it provides the necessary reciruclation - crossflow. At a given TMP you will have a specific flux. In order to avoid the fouling the system needs to be operated in a controlled fashion where the flux in controlled by the second pump. While seetting the pump on the permeate side one has to determine the ideal flow. This flow should be less than the maximum permeate flow at that particular TMP. This will give a slight backpressure on the membrane. Make sure that this back pressure is not higher than the acceptable backpressure on the membrane. Never throttle the permeate and retentate valves. For a biologics an ideal permeate flowrate would be 20 - 30 % of the feed flowrate. Due to the backpressure membrane fouling will be reduced significantly and the flux would be constant.


I hope this helps.

 

[FiltrationEngineer]

It looks like you have a feed and bleed system. If you give more details like the type of process, membranes, and specifically what problems you are having I'm sure I, or someone else can help.

 

[mechprocess]

Roger that on the feed and bleed.

I am thinking that the feed into the recirculation loop should be a variable flow rate to make up the difference between the bleed and the permeate flow rate. Our feed and bleed is set up to maintain a maximum suspended solids concentration in the re-circulation loop. The bleed returns to the filter feed tank until the suspended solids concentration in the tank has reached a targeted amount.

In my application, the permeate flow must be controlled and is a function of the Trans-membrane pressure and to a minor degree, the cross flow velocity in the filter tube media which is primarily set up to prevent fouling. We are using a filter tube with the slurry on the inside of the tubes.

I believe that we can control the permeate flow rate by controlling the TMP. I believe that this could be done using the bleed back control valve which would respond to three pressure transmitters. One on the inlet to the filter P1, another on the outlet of the filter P2 and the last on the permeate side of the filter P3. The TMP is calculated as (P1 + P2)/2 minus P3.

I believe that the flow balance can be achieved by measuring the permeate flow and the feed flow and using those values to control the speed of the feed pump. Since increasing the speed of the feed pump will add more pressure to the TMP. This will be compensated by the bleed back control valve which should operate independently from the flow balance control loop that I stated.

I don’t believe that we should modulate the permeate flow. This would provide too much feed back into the system and it would be uncontrollable.

Keep it coming.

Thanks

 

[FiltrationEngineer]

I agree that you shouldn't modulate the permeate control valve and the feed pump. It would be a mess.

I would try to have the feed pump on at 100% all the time, which would give you maximum pressure on the feed side, and modulate the permeate control valve to give a steady permeate flow (assuming you have a centrifugal pump). If you operate in this manner, you will have the maximum available TMP all the time and you use what you need by modulating the permeate valve. Then you could try and reduce pressure in your circulation loop by reducing the feed pump speed if you wanted to save some energy, but I wouldn't modulate the pump.

Trying to adjust permeate flow with TMP would probably work too. I haven't done this before, but it sounds like an interesting idea. I guess you just leave the permeate valve all the way open and run a PID loop on feed pump speed with permeate flow rate. I'd like to know how it turns out.