MF/UF membrane filtration and constant flux

[mdt1585]

I have designed the small module of MF hollow fibre membranes to test the mixture of pretreated protein solutions. The aim is just to see how quickly those membranes are fouled, however, my design for constant flux filtration is not quite right. I'm using low flow FMI pump, and pressure affects the flow a lot.

I'm thinking to doing backpressure to compensate, but I'd like to know has anyone had better idea for my set up?

 

[Lan123]

Can you provide a schematic? Is this dead end or crossflow mode?

 

[mdt1585]

Thank you very much for the reply. I have attached the schematic diagram in picture format. In the diagram, no. 1 to 5 corresponds to the solenoid valves which are controlled by the program. I'm running as filtration with regular backwashing. Every 30 minutes of filtration, it was followed by backwashing and air wetting. I'm running it as a dead-end mode. So, the only indication of the fouling is the increment in feed pressure since I have Pressure transducer after the feed valve no. 1. I normally run the system for 3-4 days to investigate the fouling rate.

As I have mentioned, I think my set up is not quite right. I collected the permeate volume every 24 hours, and then I calculated the flux based on the cumulative filtration time. I also recorded the initial flux at the very beginning of the run. I always found that as feed pressure increases within 24 hours, there's a decline in flux. Flux values are more or less the same after 24 hours. Would you have any idea for my set up to be improved?

 

[Lan123]

I don't see the attachment

 

[mdt1585]

I have attached the link to the diagram. Thanks!

 

[Lan123]

OAKAY. So what are your independant and dependant parameters. What independant paramters are you trying to hold constant

I guess if you are monitoring fouling, your dependant should be flux?

 

[membranedr]

Constant pressure in membrane technology is a complicated matter.
Even when you use a back pressure on the membrane, the pressure differential over the concentrate will still create a pressure difference at various places in the membrane module, hence there is no constant transmembrane pressure.
If the crossflow velocity is not enough different fouling will happen in the membrane which will than affect the separation and flux and so on, I guess one can understand the escalation. To obtain a real constant delta P transmembrane is a complicated process.
I have calculated the capacity improvements and they are substantial.
It requires however a new approach to the membrane module designs. This will come at an additional cost and not all membrane configurations can be used for this type of approach. What applications are you thinking off?

 

[ports394]

What you're seeing is pore fouling. That's why your fluxes decline over time, and the back washing isn't working to remove things that are lodged in the membranes pores.

 

[membranedr]

Pore fouling is a direct result of a too low crossflow velocity. When the crossflow velocity is too low, macro-molucules have a tendency to be forced into the pores. If the delta PTM get higher, the crossflow velocity needs to be increased to prevent this action. This is why in MF there needs to be a balance between velocity and pressure. If a crossflow is increased, the delta P over the concentrate is increased hence the delta PTM. For instance, MF in spirals cannot work efficiently. The required crossflow velocity would increase the delta P in the concentrate channels so high that because of this increased pressure, the polarization and mass transfer layer will effect the performance of the MF membrane and changes certain areas into a UF membrane because of surface fouling. Tubular membranes are the membranes for MF, or maybe rotating discs, anything to keep the velocity high enough