Recovery in crossflow membrane MF/UF (permeate:feed)

Hi!

What is a typical ratio of permeate to feed flow in micro-/ultrafiltration with tubular membranes (crossflow 1-3m/s)? So, I am actually looking for the ratio of retentate:permeate or the actual inlet flow circulated in the modules. It will for sure vary from application to application, but I'm interested in different applications anyway. 

In some research papers I read extremely low ratios (10-15%), but I don't think that is realistic in real world....

Taxonomy

• Treatment
• Ultrafiltration
• Filtration
• Membranes
• Industrial Water Treatment
• Membrane Filtration
• Operations

5 Answers

1. The ratio of reject to permeate might get as low as 10% if the crossflow is over a filter annulus in a rotating disk, with radial recirculation of reject. See http://www.freepatentsonline.com/7757866.html
   

   
   

2. Johanna,

   From cases I've seen and modeled, one should expect something around 85% if I'm reading your question correctly.

3. Generally, recovery is the permeate flow rate divided by feed water flow rate. However, for cross flow dead-end filtration the recovery will be difference between product volume and back wash volume divided by product volume.

   Hence, the research papers showing low ratios of about 15% shall be possible, for small cycle time as well as high backwash water quantity.

4. Bruce, thanks for the link. Very interesting!
   

   However, they don't really give the number I am looking for. They calculate the feed flow by permeate+backwash flow. But I assume the actual intake flow is higher. I don't see that number or the retentate flow in the reports I read...

5. The ETV program issued about 10 field reports on performance of membranes. The data is in the reports for calculation of recovery and ratios you are looking for. Here is the web site to access these free reports: http://archive.epa.gov/nrmrl/archive-etv/web/html/vt-dws.html#artfilt

   I hope this helps.

   Best,

   Bruce