NEP fails to start with ERROR = [Socket bind failed - errno 98 Address already in use]

I've seen an issue where port usage is not exactly as expected causing some ASAP processes to fail to start.
 
It is extremely rare but here is an example.    Can you suggest possible causes?

Thanks,
Skip

Upon ASAP startup, a NEP fails  (the specific NEP varies, it is NEPGT5 in this example)

Application [NEPGT5] is running
Server SRP_T5: Operating System Error:
Open Server Error # [16240] Severity [10] State [0]
Operating System Error # [98] Operating System Error Text [Socket bind failed - errno 98 Address already in use]

From the interfaces file:

NEPGT5
    master tcp ether zlt05140.vci.att.com 31186
    query tcp ether zlt05140.vci.att.com 31186