As a general rule, I would multiply the weight of the rocket by the amount of G forces expected when the parts snap to the end of the cord. 150 is usually a good number.
There are some things that can be done to reduce the shock experienced.
Using long shock cords can reduce the shock because the separating parts slow down before reaching the end of the cords. This works opposite on shock cord for main deploy on dual deploy systems. If the cord is too long, the chute opens but the rocket continues to fall, increasing the shock when it reaches the end of the cord.
I like to loop my shock cords back on themselves and tape every 3- 5 loops together. By the time I'm done I have a taped bundle of shock cord. It's easy to handle, doesn't tangle as easily, and has to break multiple layers of tape to deploy, which helps reduce the shock on the cord.
Using electronic deployment helps too. It ensures the deployment event occurs at apogee. As long as the flight was vertical, that should be the point with a safe, low air speed.