Dear Reetu,
Without more information, the first thing I can suggest is to verify that your geometries and calculation parameters (charge, multiplicity, etc.) are sensible.
If you could post the output file, members of the forum would have more information to work with.
But for a quick shot in the dark, you could try running the calculation with like this MKL_CBWR=COMPATIBLE molpro instead of just molpro
If your installation of Molpro has been built with the MKL linear algebra library, this will force it to take a different, slower code path that may be slightly more numerically accurate, at the cost of speed. This can sometimes nudge problematic calculations out of hitting a singularity.
Alternatively you could also try tightening various integration thresholds to nudge the calculation, like this:
gthresh,twoint=1.d-14
gthresh,compress=1.d-14
gthresh,zero=1.d-14
Of course this also has a speed and memory/disk usage penalty.
Best,
Tibor