I've corrected it and it worked.

Please pay attention to this part of your file:

  I    1    (CENTRE  1)  -8.79274757 -1.79272340  0.00000000  CHARGE = 53.0
  I    2    (CENTRE  2)  -6.32872640  2.73017867  0.00000000  CHARGE = 53.0

When ECP is used, the value after CHARGE should be number of explicitly represented valence electrons rather than element index. Please manually modify it according to the ECP you used.

Gaussian doesn't have this problem.

My problem occurs when I use effective core potentials for relatively large atoms like Br and I. In this case I get unreasonable results where all the molecular surface has a positive electrostatic potential value.

I get the effective core potentials from the basis-set exchange site: https://www.basissetexchange.org/. When I was working with Gaussian09, it was okay and the results were sound, but this is the first time where I use GAMESS with explicit basis sets.

For instance, for iodine (I2), I get these results from Multiwfn 3.7 (GAMESS input and output files are attached at the end of the post):

Volume:   710.01669 Bohr^3  ( 105.21364 Angstrom^3)
Estimated density according to mass and volume (M/V):    4.0058 g/cm^3
Minimal value:   5880.60843 kcal/mol   Maximal value:   7195.40597 kcal/mol
Overall surface area:         407.16839 Bohr^2  ( 114.01878 Angstrom^2)
Positive surface area:        407.16839 Bohr^2  ( 114.01878 Angstrom^2)
Negative surface area:          0.00000 Bohr^2  (   0.00000 Angstrom^2)
Overall average value:    9.99750281 a.u. (   6273.53299 kcal/mol)
Positive average value:   9.99750281 a.u. (   6273.53299 kcal/mol)
Negative average value:          NaN a.u. (          NaN kcal/mol)
Overall variance (sigma^2_tot):  0.43247968 a.u.^2 (170297.00593 (kcal/mol)^2)
Positive variance:        0.43247968 a.u.^2 ( 170297.00593 (kcal/mol)^2)
Negative variance:        0.00000000 a.u.^2 (      0.00000 (kcal/mol)^2)
Balance of charges (nu):   0.00000000
Product of sigma^2_tot and nu:   0.00000000 a.u.^2 (    0.00000 (kcal/mol)^2)
Internal charge separation (Pi):   0.55515652 a.u. (    348.36627 kcal/mol)
Molecular polarity index: 272.04588592 eV (   6273.53299 kcal/mol)
Nonpolar surface area (|ESP| <= 10 kcal/mol):      0.00 Angstrom^2  (  0.00 %)
Polar surface area (|ESP| > 10 kcal/mol):        114.02 Angstrom^2  (100.00 %)

The GAMESS input and output (extracted wfn file) are attached to these links:

I2-input.txt

I2.wfn