It sounds like you didn't update the server.conf to update the URLs for your sandbox instance. Since the kStop script read the server.conf file to find the URL it needs to send the shutdown command to, you need to make sure that the server.conf file has the correct URL. If server.conf file for both the Prod and Sandbox instances are using the same URL, it's behaving exactly as it should.
You also need to change all of the multicast ports used by the clone instance so it does not communicate with the primary instance. And you need to run kUpdateHtml.sh after making any changes to server.conf.
Though the parameters line RMI_URL, KINTANA_SERVER, multicast ports were modified per the new instance (Clone) there were not effective since I was unable to stop the services (Whenever I stopped the services on the Clone it stopped PROD services too)
This is what I did to cut the connection between Clone & PROD
1. Stop the servers (PROD & CLONE) off hours
2. Modify each entry related to the clone instance on the server.conf manually (Not by running kConfig.sh)
3. Run kUpdateHtml.sh
4. Clear tmp & work folders
5. Restart the services on PROD first and then Clone.
This way the base URL and other parameters are now different and the connection is broken between Clone server & PROD.