Troubleshooting Cake Filtration Experiments

Constant pressure tests are the most frequently used method of obtaining laboratory data for filter performance and scale-up as experimental arrangements are relatively straightforward.  It is conventional practice to record cumulative time and volume of filtrate and convert these into a plot of time/volume vs. volume.  Ideally a straight line is produced from which cake resistance and filter medium resistance can be calculated.  In reality, however, deviations from this ideal are frequently observed as shown in the figure.

Non-linearities close to the start of filtration are normally due to interactions between the filter medium and the initial layers of particles forming the cake and/or poor drainage characteristics of the test filter. 

Non-linearities toward the end of filtration can occur due to:

• If the filtration is carried out in a filter with a fixed volume chamber, then when the chamber is full of cake the filtrate volume collected in any time interval is reduced
• If a cake is formed and the feed slurry is completely used, then the applied pressure difference serves to deliquor the cake and again the filtrate flow rate is considerably reduced. 
• If settling of the solids has occurred during the test, then the data sequence tends
  to assume a horizontal plateau as permeation (rather than filtration) proceeds.

Any of the above deviations from linearity can lead to significant errors when calculating cake and filter medium resistances as well as final cake moisture.  To promote better data analysis and more reliable scale-up it is recommended that:

1. Careful observations are made during an experiment, particularly toward the end so that the final cake condition is known
2. Suitable corrections are applied to account for the loss of liquid from the final cake due to any deliquoring
3. The linear region of the time/volume vs. volume is chosen carefully.

 

 

© 2005 American Filtration & Separations Society All rights reserved