The case can not run if refine the mesh

I am conducting research on the adhesion and rebound of small particles colliding with large particles. When the grid is set to 30 * 30 * 1 (the grid size is about 5e-6), it can calculate. However, a portion of the computational domain extends into the interior of large particles (wall conditions), and it is unknown what is wrong. the Fig is shown as belowed.

Therefore, an attempt was made to refine the grid 1001002(the mesh size is about 2e-6), there are no errors reported but the simulation wont run.
MFIX WITH CARTESIAN GRID IMPLEMENTATION.
RE-INDEXING IS TURNED OFF.
INITIALIZING VELOCITY NODES…
ESTIMATING POTENTIAL SCALAR CUT CELLS…
INTERSECTING GEOMETRY WITH SCALAR CELLS…
Warning: UNABLE TO PROPAGATE F_AT ARRAY FROM myPE=. 0
THIS USUALLY INDICATE A RANK WITH NO ACTIVE CELL
YOU MAY NEED TO ADJUST THE GRID PARTITIONING
TO GET BETTER LOAD_BALANCE.
INTERSECTING GEOMETRY WITH U-MOMENTUM CELLS…
Warning: UNABLE TO PROPAGATE F_AT ARRAY FROM myPE=. 0
THIS USUALLY INDICATE A RANK WITH NO ACTIVE CELL
YOU MAY NEED TO ADJUST THE GRID PARTITIONING
TO GET BETTER LOAD_BALANCE.
INTERSECTING GEOMETRY WITH V-MOMENTUM CELLS…
Warning: UNABLE TO PROPAGATE F_AT ARRAY FROM myPE=. 0
THIS USUALLY INDICATE A RANK WITH NO ACTIVE CELL
YOU MAY NEED TO ADJUST THE GRID PARTITIONING
TO GET BETTER LOAD_BALANCE.
INTERSECTING GEOMETRY WITH W-MOMENTUM CELLS…
Warning: UNABLE TO PROPAGATE F_AT ARRAY FROM myPE=. 0
THIS USUALLY INDICATE A RANK WITH NO ACTIVE CELL
YOU MAY NEED TO ADJUST THE GRID PARTITIONING
Reading mesh stats: /home/gpf/mfix/try.10.14/MESH_STATS.LOG

The case is attached. Can you show me how to resolve it.
try.10.14.zip (9.2 MB)