Matthew,
Thanks for the comments,
I would like to confirm the following:-
1. The errors taken into calculations are independent of each other.
2. It is understood that there would only be a probability of n/3 errors occuring in 50% of the cases and 2n/3 in 93% of the cases.
Given the above statements, it would imply that if I have a max error of 2% in system, for 50% of the cases I may expect a probability of 0.67% error, for 93% of the cases 1.34% error and so on (in accordance with the CEP definition stated in the original post)
Thats what I finally intended to confirm.
Thanks