Problem with large time scale and minute length scale
 

Hi all,
I'm simulating flow in a porous media, using solver buoyantBoussinesqPimpleFoam with porous resistance added to NS equations. The Kolmogorov Time scale for my problem is around 200 seconds, and the length scale is around 1 micrometer. The total time for observation is 40000 seconds, and my domain size is 10.2 cm * 5.1 cm * 0.051 cm.
How important is the Courant Number limit for PIMPLE algorithms? If it is important, how do I avoid violating it with such a large time-step and such a minute mesh-cell length? Is there a work-around for this problem? What should be my time-step and mesh-size?
Thanks in advance