Some Basic Definitions

Definitions of
Key Statistical Parameters
Statistical Distributions

• Definition of probability
• Probability density functions
• Types of distributions
• Normal distribution
• Cumulative distribution functions
 Frequency of Occurrence
Frequency of occurrence of a given event is the number of
times this event exists within the population of the data set

Example 1:
Frequency of occurrence of blue balls = 8
Frequency of occurrence of red balls = 10
Frequency of occurrence of yellow balls = 12

Example 2:
Frequency of occurrence of any number
from 1 to 6 on each dice = 6
 Definition of Probability
Probability is the relative frequency (or chance) of
occurrence of a given event among a population of events

Example 1:
P {drawing a blue ball} = 8/30
P {drawing a red ball} = 10/30
P {drawing a blue ball then a red ball} = 80/900

Example 2:
P {getting a 3 on dice A} = 1/6
P {getting a 5 on dice B} = 1/6
P {getting a 3 on dice A and 5 on dice B} = 1/36
 Frequency for Variables
Consider variable X that has values in the range:

Xmin ≤ X ≤ Xmax

This range can be divided into a number of intervals and the

number of occurrences within each interval recorded as:

Interval 1 2 i-1 i i+1 n

Frequency f1 f1 fi-1 fi fi+1 fn

Relative frequency for each interval can be expressed as:

n

if = f /∑ f
r
i i
1
 Frequency for Variables
Example:
A given core permeability data set is:
473 231 253 326 534 347 129 442 609 338
275 366 189 574 725 439 249 634 844 386

Using an interval of 100 md:

Interval 100-200 200-300 300-400 400-500 500-600 600-700 700-800 800-900

Frequency 2 4 5 3 2 2 1 1

Rel. frequency 0.1 0.2 0.25 0.15 0.10 0.10 0.05 0.05
 Frequency Histogram
Frequency and or relative frequency can be presented in the form
of histogram for discrete variables (minimum of 7 intervals)
For the permeability example:

0.3

Relative 0.2
Frequency

0.1

0
0 100 200 300 400 500 600 700 800 900
Permeability, md
 Probability Density Function
When the relative frequency histogram is replaced by a
continuous curve (very small intervals), the resulting curve
represents the probability density function PDF

PDF
or
Relative
Frequency

Variable
 Probability Density Function

Total area under curve = 1

PDF
or
P(a<X<b)
Relative
= Area
Frequency

Xmin a b Variable X Xmax

It follows that:
P{X < Xmin}= P{X > Xmax} = 0
P{X > Xmin} = P{X < Xmax} = P{Xmin < X < Xmax} = 1
 Types of Statistical Distributions

PDF PDF

Variable Variable

Uniform Distribution Normal Distribution

 Types of Statistical Distributions

PDF PDF

Variable Variable

Symmetrical Triangular Distribution Skewed Triangular

Distribution
 Types of Statistical Distributions

PDF PDF

Variable Variable

Log-Normal Distribution Root-Normal

Distribution
 The Normal Distribution

0.5
PDF
Mean = 200 0.4
Std Dev = 30
0.3

0.2

0.1

0
-4 -3 -2 -1 0 1 2 3 4
Standard deviations
80 110 140 170 200 230 260 290 320
Real Variable Values
 The Normal Distribution
1 − 0 .5 x 2
• The normal distribution PDF is: PDF = e
• Probability is tabulated as: 2π
X Standard Deviations P[Variable is within ±xσ from mean]
0 0
0.25 0.198
0.50 0.383
0.75 0.547
1.00 0.682
1.25 0.789
1.50 0.866
1.75 0.920
2.00 0.954
2.50 0.988
3.00 0.997
3.50 0.999
4.00 1.000
 The Normal Distribution
Example
A stratal element has:
Mean thickness = 5 m Standard deviation σ = 2 m
Range = 0.15 – 12 m

1. P{2 ≤ thickness ≤ 8} = P{x = 3/2} = 0.866 or 86.6%

2. P{4 ≤ thickness ≤ 7} = 0.5[P{x = (5-4)/2} + P{x = (7-5)/2}]

= 0.5[0.383 + 0.682] = 0.533 or 53.3%

3. P{thickness = 5} = 50%
P{thickness ≤ h} = 90% means (1/2)P{x = (h - 5)/ σ} = 40%
By interpolation: P{1.29} = 80%
hence; (h - 5)/2 = 1.29 h = 7.6 m
or P{thickness ≤ 7.6 m} = 90%
 Cumulative Distribution Function
• The CDF is the integration of the distribution PDF
• CDF(x) = probability that the variable value is ≤ x

CDF CDF
PDF PDF

Variable Variable
Positive Skewness Negative Skewness
 Types of Averages
Measurement number 1 2 3 ------------------- n-1 n
Measured value X1 X2 X3 Xn-1 Xn

_ n
1
Arithmetic Average X ar = ∑ Xi
n 1
1n
_ ∑ log( X i )
Geometric Average X geom = 10 n1

_
n
Harmonic Average X har = n
1
∑1 Xi
 Using Different Averages for Permeability

This represents parallel fluid flow h1 k1

Use arithmetic average h2 k2

Avg. k = Σ(kihi)/ Σhi

hn kn

h1 k1
This represents series fluid flow
h2 k2
Use harmonic average
Avg. k = Σhi / Σ(hi/ki) hn kn
Using Different Averages for Permeability

h1 k1 h2 k2

h3 k3

hn kn

This represents random combination of series

and parallel fluid flow
Use geometric average
log(Avg. k) = Σ[hi log(ki)] / Σhi
Surface and Reservoir Water Cuts

qw Surface m3/day
qo STm3/day

Bwqw Rm3/day
Boqo Rm3/day

Surface water cut fws = qw / (qw + qo)

Reservoir water cut fwr = Bwqw / (Bwqw +Boqo)

Hence; fwr = fws / [fws + (1 – fws)(Bo/Bw)]

Producing and Solution Gas-Oil Ratio

qg = Rsqo + qgf Sm3/day

qo STm3/day

Bgqgf Rm3/day Solution GOR

Rsqo Sm3/day Rs Sm3/STm3

Producing GOR Rp = qg / qo

Hence; qgf = qg - Rsqo