Page 92 - 6685

n
 nk i i
k   i 1 n , (5.1)
n i
 i 1
where k i, n i taken from permeability distribution polygon;
- variance (mean-square deviation) defines the fluctuations
of the actual value of the parameter near the mean in absolute units
n
2
   kk i  n i
    i 1 n , (5.2)
 n i
 i 1
- coefficient of variation, which is ratio of dispersion to
mathematical expectation

    , (5.3)
k
shows fluctuations of the parameter in decimal quantity from the
average.
Considering the fact that not all of the possible research
activities could be done the average values of parameters could be
estimated with errors (non-sampling errors). Non-sampling errors
could be estimated as:
t 
   , (5.4)
N
where - limiting sampling error; - variance; N - number of
samples; t - coefficient with value depended from given
probability (t=1 when probability 0.68; t=2 when probability 0.85;
t=3 when probability reaches 0.97);
In case when the error value is set by the regulation ( it
should not exceed the pre-set value), in order to prevent error
value overrunning the established limits at a predetermined
probability, by using equation (5.4), determine the number of
models that need to be analyzed:

92