I have run 10 instances of the process eg, process name is BG

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 1 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>a.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 2 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>b.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 3 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>c.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 4 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>d.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 5 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>1.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 6 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>2.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 7 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>22.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 8 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>3.txt &

nohup /WP01IRB1_irbapp/IRBWPROD/RB/bin/BG -c 9 -t 23 -a '-caTop TESTBILLCYCLE='5FEB13_81PT19NPT''>4.txt &

Based on what you presented there is no way to tell what is going on. I am guessing you are running cycle billing, and that you divided accounts from one cycle or group into sub-cycles.

As a wild guess, did you flatten out the the number of accounts in each sub-cycle, e.g., do they all contain approximately the same number of accounts, e.g. say 500 accounts?

yes i have set this bill cycle for 5 lac accounts .

So you are saying 500K accounts are there is each cycle. But if this is cycle billing then you have to consider the account total usages count as well. May be the total usages are much higher compared to the instances that are getting completed early.

If your process BG is connecting to DB, then I suggest you to monitor the SQL sessions and see at what stage it is taking time. Like Jim mentioned it is really difficult to tell what exactly is the reason behind because we have no idea about your process BG and what volume of data it is dealing with!