voronoi occupation

Dear all,
I am using LAMMPS 2014.07.14 for radiation damage simulation. when I use “voronoi occupation” command for simulation of radiation damages produced in a Fe bcc single crystal, I have found some following points and have some questions about them:
1- some zeros for the first column of compute output and some more than 1 numbers for the second column are displayed even before radiation occurred (before given more kinetic energy to a single atom in the crystal). It seems that thermal fluctuation itself can produce some vacancy and interstitials. If it is correct how can I distinct between vacancy and interstitials produced by radiation or thermal fluctuation?
2- number of vacancies (zeros in first column) is not equal to the total number of interstitials in any timestep, but I expect to have equal number of vacancies and interstitials in single pure crystal.
3- If a number in column 2 is more than 1 (as an example 3) it means that the number of interstitials in the voronoi cell is equal to the number minus 1 (equal to 2 for the example)? Should I sum over any number of second column x minus one (sum over (x-1)) to calculate the total number of the interstitials?

Sincerely yours,
S.M.Modarresi

Also, This is my input:

This simulation is done to examine the VORONOI/Atom command for Vacancy Interestitial determination 4-5-93

dimension 3
boundary p p p

units metal
atom_style atomic

create box

region crystal block -37 37 -37 37 -17 17 units box

747434 angstrom

create_box 1 crystal

lattice bcc 2.855324

create_atoms 1 region crystal
group crystal region crystal

mass 1 55.9

pair_style eam/fs

pair_coeff * * fe_2.eam.fs Fe

velocity crystal create 300 4928459 mom yes rot yes dist uniform
fix 1 crystal nvt temp 300 300 100

compute iv all voronoi/atom occupation

thermo 1000
timestep 0.001

dump 1 all xyz 1000 dump.xyz
dump 2 all movie 500 movie.mpg element element axes yes 0.8 0.02 view 60 -30
dump 3 all custom 10000 custom.dmp c_iv[1] c_iv[2]

run 30000

primary knock on atom definition

region pka sphere 35.6915 1.42 1.42 1.45 units box
group pka region pka

fix 2 pka nve

velocity pka set -590 NULL NULL units box

timestep 0.0001

dump_modify 3 every 5000
dump 4 all xyz 10000 dump1.xyz
dump 5 pka custom 1000 custom1.dmp c_iv[1] c_iv[2]

run 100000

Perhaps Daniel can comment.

Steve