where: e = expected error in calculating the sample mean 



z = standard normal distribution value associated with a 

 confidence coefficient of (l-a/2) (see Appendix A) 



a = desired confidence indicator 



Solving for n: 



-,2 



(4) 



(5) 



If the expected error is given in terms of a percentage of the mean 

 measurement value, then: 



e - c 



(6) 



where c is the desired accuracy of the estimate in terms of a percentage 

 of the mean value (x) . 



Substituting into Equation 5, 



The term s /x is also designated the Coefficient of Variation, 

 denoted by 6 . Therefore, 



— 6 

 c x 



(7) 



The preceding discussion was developed for a sample taken from an 

 infinite population. When the population size is finite, as in the case 

 of a waterfront facility, the equation for the number of required samples 

 can be modified to include a term accounting for the population size as 

 follows : 



. z 



" (N - 1) 1_ 



2 . 2 „ N 



-1 



(8) 



where N is the total population size. 



As the population size (N) gets larger, Equation 8 reduces to that 

 shown in Equation 7. 



