Number of CPUs is not enough when submitting to queue

Question

I am trying to submit a job that uses 120 processes to a queue system. In this cluster, each node has 40 CPUs, so I am requesting 3 nodes with 40 processes on each. My PBS scrip