Comments on statistical error in ORNL's program NMTC