PET176

Reservoir Rock Properties

Prepared by
Dr. Turhan Yildiz and Dr. Erdal Ozkan
6.2.1 Klinkenberg Effect at low pressures
Limitations of Darcy’s Law

 Typically, we measure permeability by flowing gas through core plugs at

low pressures
2 𝑞𝑠𝑐 𝜇 𝑝𝑠𝑐 𝐿
 𝑘𝑎𝑔 =
𝐴 𝑝12 −𝑝22
 Permeability calculated using the eq. above at low test pressures is
higher than the true permeability
 This is due to a physical phenomenon known as gas slippage at the pore
surfaces
 Gas slippage phenomenon in pores is formulated by Klinkenberg

4
Gas Flow at Low Pressures and Klinkenberg Effect

 Klinkenberg developed a theory explaining the gas slippage

 When a gas is flowing along a solid wall, the layer of gas next to the
surface is in motion with respect to the solid surface
 If the wall has a zero velocity then the velocity of the gas layer in the
immediate vicinity of the wall has a finite value
 As a consequence the quantity of gas flowing through a capillary is
larger than would be expected from Poiseuille’s formula
 Klinkenberg has proven that the permeability to gas and liquid are
related through the expression below

5
Gas Flow at Low Pressures and Klinkenberg Effect

 𝑘𝑎𝑔 = 𝑘𝐿 1 + 𝑏𝐾 /𝑝 = 𝑘𝐿 + 𝑏𝐾 𝑘𝐿 /𝑝
 kag = apparent gas permeability, md
 kL = liquid permeability, md
 bK = Klinkenberg slip coefficient, psia
 At high pressures, the second term containing the Klinkenberg slip
coefficient tends to zero; gas and liquid permeabilities become same
 At low pressures, gas permeability is higher than liquid permeability
 Gas permeability at low pressures is also a function of average pressure

6
Gas Flow at Low Pressures and Klinkenberg Effect

 As mean test pressure decreases, apparent gas permeability increases

 Klinkenberg slip coefficient is a function of gas type as well as porous
rock itself
 To determine intrinsic (true or liquid) permeability, we measure
apparent gas permeability at several mean pore pressure
 𝑘𝑎𝑔 = 𝑘𝐿 + 𝑏𝐾 𝑘𝐿 /𝑝
 Notice that the eq. above has the form of a straight line eq.
 This observation may be used to predict both intrinsic permeability and
Klinkenberg slip coefficient as outlined below

7
Gas Flow at Low Pressures and Klinkenberg Effect

 Step 1: Starting with the lowest mean pressure, measure apparent gas
permeability at several mean pressures, typically four tests
 Step 2: Construct a Cartesian plot of kag versus 1/𝑝
 kag versus 1/𝑝 plot is referred to as Klinkenberg plot
 The Klinkenberg plot should display a straight line
 Step 3: Liquid permeability kL is equal to intercept of the straight line
 Step 4: The Klinkenberg coefficient may be computed from the slope of
the straight line on the Klinkenberg plot

8
Klinkenberg Plot

kag

kL

0 1/p~

9
Gas Flow at Low Pressures and Klinkenberg Effect

 𝑘𝑎𝑔 = 𝑘𝐿 + 𝑏𝐾 𝑘𝐿 /𝑝
 kag versus 1/𝑝 plot
 Compute slope and intercept
 y0K = intercept of the Klinkenberg plot
 mK = slope of the Klinkenberg plot

 𝑘𝐿 = 𝑦0𝐾
 𝑏𝐾 = 𝑚𝐾 /𝑘𝐿 = 𝑚𝐾 /𝑦0𝐾

10
Klinkenberg Slip Effect – Example 5

 Klinkenberg presented extensive data showing the relationship

between single-phase gas and liquid permeabilities
 Test data reported for Core Sample A
 kL of Core Sample A using liquid is 23.66 md
 Apparent gas permeability and mean pressure data extracted from
Klinkenberg’s work are listed in Table 3
 Construct the Klinkenberg plot
 Predict intrinsic permeability and Klinkenberg slip coefficient

11
Klinkenberg Slip Effect – Example 5
Table 3 – Apparent gas permeability at low pressures (After Klinkenberg)
p~ 1/p~ kag p~ 1/p~ kag
(psia) (1/psia) (md) (psia) (1/psia) (md)

0.1523 6.56812 290.00 44.1909 0.02263 25.43

0.1597 6.25996 280.00 47.3799 0.02111 25.09
0.1747 5.72294 259.00 64.1627 0.01559 25.02
0.2400 4.16692 197.00 74.2001 0.01348 24.72
0.2973 3.36360 165.00 74.9496 0.01334 24.62
0.3433 2.91292 148.00 93.1726 0.01073 24.34
0.4628 2.16087 117.00 93.1726 0.01073 24.29
0.8702 1.14923 75.60 94.4953 0.01058 24.25
1.3053 0.76611 60.00 131.3822 0.00761 23.79
2.1515 0.46479 46.60 131.3822 0.00761 23.85
5.0451 0.19821 34.20 135.0562 0.00740 23.93
11.7421 0.08516 28.60 182.9652 0.00547 23.66
23.0727 0.04334 26.79 183.4061 0.00545 23.82
23.2344 0.04304 26.58 241.6022 0.00414 23.67
26.4822 0.03776 26.49 241.7492 0.00414 23.77
29.9358 0.03340 26.23 294.0670 0.00340 23.55
30.0533 0.03327 26.33 294.0670 0.00340 23.65
39.0473 0.02561 25.22

12
Klinkenberg Slip Effect – Example 5

13
Klinkenberg Plot – Example 5

14
Klinkenberg Plot – Example 5

 𝑘𝑎𝑔 = 𝑘𝐿 + 𝑏𝐾 𝑘𝐿 1/𝑝 = 24.78 + 41.02 1/𝑝

 𝑦0𝐾 = 24.78
 𝑚𝐾 = 41.02

 𝑘𝐿 = 𝑦0𝐾 = 24.78 md
 𝑏𝐾 = 𝑚𝐾 /𝑘𝐿 = 41.02 /24.78 = 1.66 psia
 The 𝑘𝐿 = 24.78 md predicted from gas flow tests agrees very well with
the liquid permeability of 𝑘𝐿 = 23.66 md from liquid flow tests

15
Effect of Gas Composition on Klinkenberg Slip

 Apparent gas permeability measured at low pressures is influenced by

mean test pressure
 Experimental studies using different gases such as nitrogen, carbon
dioxide, air, hydrogen, and helium
 Apparent gas permeability is also controlled by gas type
 Type of gas also controls numerical value of Klinkenberg slip coefficient
 Next figure illustrates apparent permeability to different gases

16
Effect of Gas Composition on Klinkenberg Slip

kag
Gas 1

Gas 2

Gas 3
kL

0 1/p~

17
Effect of Gas Composition on Klinkenberg Slip

 Different gases exhibit different slopes on Klinkenberg plot

 This implies that Klinkenberg slip coefficient is influenced by gas
composition

18
Gas Composition Effect on Klinkenberg Slip – Example 6

 kag to helium, hydrogen, and nitrogen measured as a function 𝑝

 L = 3.25 cm and d = 1.89 cm
 Test data taken from McClure’s thesis are given in Table 4
 𝑝 varies from 22.5 psia to 82.5 psia
 kag values computed using 𝑘𝑎𝑔 = 2 𝑞𝑠𝑐 𝜇 𝑝𝑠𝑐 𝐿/𝐴 𝑝12 − 𝑝22
 Measured apparent gas permeabilities are tabulated for each gas
 Construct Klinkenberg plot for each gas
 Estimate liquid permeability and Klinkenberg slip coefficient

19
Gas Composition Effect on Klinkenberg Slip – Example 6
Table 4 – Experimental data for three different gases
Helium Hydrogen Nitrogen
p~ 1/p~ kag p~ 1/p~ kag p~ 1/p~ kag
(psia) (1/psia) (md) (psia) (1/psia) (md) (psia) (1/psia) (md)
22.519 0.0444 6.795 22.474 0.0445 6.132 22.509 0.0444 5.593
27.508 0.0364 6.490 27.474 0.0364 5.897 27.495 0.0364 5.442
32.476 0.0308 6.238 32.463 0.0308 5.737 32.513 0.0308 5.300
37.475 0.0267 6.061 37.490 0.0267 5.611 37.490 0.0267 5.181
42.473 0.0235 5.917 42.486 0.0235 5.522 42.474 0.0235 5.138
47.487 0.0211 5.893 47.468 0.0211 5.447 47.483 0.0211 5.077
52.491 0.0191 5.779 52.467 0.0191 5.365 52.467 0.0191 5.029
57.469 0.0174 5.690 57.474 0.0174 5.308 57.451 0.0174 4.992
62.478 0.0160 5.618 62.456 0.0160 5.249 62.456 0.0160 4.961
67.487 0.0148 5.566 67.475 0.0148 5.202 67.475 0.0148 4.925
72.495 0.0138 5.498 72.465 0.0138 5.150 72.501 0.0138 4.880
77.469 0.0129 5.468 77.470 0.0129 5.125 77.470 0.0129 4.870
82.478 0.0121 5.394 82.469 0.0121 5.107 82.469 0.0121 4.827

20
Gas Composition Effect on Klinkenberg Slip – Example 6

21
Gas Composition Effect on Klinkenberg Slip – Example 6

22
Gas Composition Effect on Klinkenberg Slip – Example 6

Table 5 – Permeability estimates using different gases

Gas mK y0K kL bK
() (md·psia) (md) (md) (psia)
Helium 42.545 4.932 4.932 8.626
Hydrogen 32.089 4.737 4.737 6.774
Nitrogen 23.473 4.572 4.572 5.134

23
Gas Composition Effect on Klinkenberg Slip – Example 6

 kL obtained from the intercept of the straight lines on Klinkenberg plot

varies slightly depending on the type of gas
 Helium test data give kL = 4.93 md
 Test data using nitrogen gas yield kL = 4.57 md
 Slopes of the straight lines on the Klinkenberg plot and the Klinkenberg
slip coefficient computed from the test data are different
 Helium gas yields bK = 8.63 psia
 Nitrogen gas has bK = 5.13 psia

24
6.2.1 Forchheimer’s Equation for high velocities
High Velocity Flow in Porous Media

 According to Darcy’s law, local pressure gradient is linearly proportional

to in-situ fluid velocity
 This could be directly observed by rearranging Darcy’s equation
𝑑𝑝 𝜇
 − = 𝑢
𝑑𝑥 𝑘
 Darcy’s law applies only when fluid flows at low velocities
 At high flow velocities, Darcy’s equation fails
 At high fluid velocities, the relationship between pressure gradient and
velocity is represented by a quadratic equation

26
High Velocity Flow in Porous Media

𝑑𝑝
 − = 𝑎 𝑢 + 𝑏 𝑢2
𝑑𝑥
 a and b are constant coefficients
 The quadratic relationship between the pressure gradient and fluid
velocity has been observed by Forchheimer in 1901
 The quadratic equation above is known as Forchheimer equation
 Next figure illustrates the quadratic relationship between pressure
gradient and fluid velocity

27
High Velocity Flow in Porous Media

Darcian flow

dp/dx

28
High Velocity Flow in Porous Media

 At low fluid velocities, Darcy’s law still applies

 At high fluid velocities, the relationship between pressure gradient and
velocity start to deviate from Darcy’s law
 Pressure gradients are higher than those predicted by Darcy’s equation
 Pressure gradient is related to velocity in a nonlinear fashion
 Later researchers have confirmed the Forchheimer’s findings
 Theoretical and experimental works have proved that a better version
of the Forchheimer’s equation containing the Darcy’s law should be as
follows

29
High Velocity Flow in Porous Media
𝑑𝑝 𝜇
 − = 𝑢 + 𝛽𝜌 𝑢 𝑢
𝑑𝑥 𝑘
 b is the coefficient of the quadratic term
 Many different names for b coefficient
 coefficient of inertial resistance, inertial coefficient
 turbulence coefficient, turbulence parameter, turbulence factor

 high-velocity coefficient

 non-Darcy flow coefficient

 b factor

 quadratic coefficient

 Forchheimer coefficient

30
High Velocity Flow in Porous Media
𝑑𝑝 𝜇
 − = 𝑢 + 𝛽𝜌 𝑢 𝑢
𝑑𝑥 𝑘
 p = pressure, atm.
 x = distance, cm
 dp/dx = pressure gradient, atm/cm
 m = viscosity, cp
 k = permeability, d
 u = local fluid velocity, cm/s
 b = Forchheimer coefficient, atm·s2/g
 r = local fluid density, g/cc

31
High Velocity Flow in Porous Media – Example 7

 Experimental data reported by Brownell et al. (1950)

 An artificial linear porous media composed of packed glass spheres
 The glass spheres had an average diameter of 0.2 inches
 Porosity of the pack was 0.412
 They measured the pressure gradient across the pack of glass spheres
by flowing water through it
 The data extracted from their study are tabulated in Table 6
 Additionally, m = 0.945 cp and r = 0.998 g/cc

32
High Velocity Flow in Porous Media – Example 7

 Demonstrate the limitations of Darcy’s law

 Fit the experimental data with a second degree polynomial
 Using Forchheimer equation, estimate
 permeability
 Forchheimer coefficient

33
High Velocity Flow in Porous Media – Example 7
Table 6 – Pressure gradient as a function of velocity
(data adapted from Brownell et al.)
Test # u m |dp/dx|
() (cm/s) (cp) (atm/cm)
0 0.00 0.00000000
72 3.41 0.955 0.00040918
73 3.93 0.955 0.00052387
74 4.61 0.955 0.00068041
75 5.15 0.955 0.00080440
76 5.99 0.956 0.00100124
77 6.82 0.956 0.00121978
78 7.69 0.953 0.00154681
79 8.49 0.952 0.00175294
80 9.49 0.955 0.00210322
81 10.47 0.930 0.00254339
82 11.20 0.931 0.00283633
83 12.18 0.931 0.00331680
84 13.85 0.931 0.00423899
85 15.61 0.931 0.00514569
86 17.37 0.931 0.00613763

34
High Velocity Flow in Porous Media – Example 7

35
High Velocity Flow in Porous Media – Example 7

 In previous figure, the solid circles represent the measured data

 Solid line is the second degree polynomial fit to the measured data
 At low flow velocities, the pressure gradient and flow velocity may
appear linearly related as predicted by the Darcy’s law.
 The measured data deviate from the linear trend described by the
Darcy’s law
 The higher the flow velocity is, the larger the deviation between the
measured data and the apparent linear trend at low velocities

36
High Velocity Flow in Porous Media – Example 7

 Using a spreadsheet program, we determine the coefficients of the

second degree polynomial fitting the measured test data
 The polynomial equation fitting the data is given below
 𝑑𝑝/𝑑𝑥 = 𝑎 𝑢 + 𝑏 𝑢2 = 6.90279 × 10−5 𝑢 + 1.65942 × 10−5 𝑢2
 𝑎 = 6.90279 × 10−5
 𝑏 = 1.65942 × 10−5
 Notice that, in the polynomial equation, the intercept value is zero

37
High Velocity Flow in Porous Media – Example 7

 𝑘 = 𝜇/𝑎 = 0.945/6.90279 × 10−5 = 13,690 d

 𝛽 = 𝑏/𝜌 = 1.65942 × 10−5 /0.998 = 1.663 × 10−5 atm ∙ s2 /g = 16.85 1/cm

38
Unit and Dimension for Forchheimer Coefficient
 Rearranging Forchheimer equation,
−𝑑𝑝/𝑑𝑥 –𝜇𝑢/𝑘
 𝛽=
𝜌𝑢 𝑢
 Substituting the units for all variables
atm/cm atm∙cm3 ∙s∙s atm∙s2
 𝛽 =𝛽 =𝛽
g/cm3 cm/s cm/s cm∙g∙cm∙cm g
 Substituting the dimensions for all variables
m/Lt2 /L m∙L3 ∙t∙t 1
 𝛽 =𝛽 =𝛽
m/L3 L/t L/t L∙t2 ∙L∙m∙L∙L L
 b dimension is the reciprocal of length

39
Unit and Dimension for Forchheimer Coefficient

 Unit of atm·s2/g for Forchheimer coefficient is somewhat unusual

 Typically Forchheimer coefficient is given in unit of either 1/cm or 1/ft
 In cgs units, unit of 1/cm
 In field units, unit of 1/ft
 The conversion factors to convert between the units of the Forchheimer
coefficient are given below

40
Unit Conversion for Forchheimer Coefficient

 1 (atm-s2/g) =1.0134x106 (1/cm)

 1 (1/cm) = 9.868x10-7 (atm·s2/g)
 1 (1/ft) = 0.032808 (1/cm)
 1 (1/cm) = 30.48 (1/ft)
 1 (atm-s2/g) = 3.089x107 (1/ft)
 1 (1/ft) = 3.238x10-8 (atm·s2/g)

41
High Velocity Flow in Porous Media

𝑑𝑝 𝜇
 − = 𝑢 + 9.868 × 10−7 𝛽𝜌 𝑢 𝑢
𝑑𝑥 𝑘
 In the eq. above, Forchheimer coefficient, b, is in the unit of 1/cm
 High velocity flow is generally encountered in gas wells
 When we try to model fluid flow into gas wells, we typically use the
field units
 Forchheimer equation in field unit is as follows

42
High Velocity Flow in Porous Media

𝑑𝑝 5 𝜇𝑔
 − = 1.5804 × 10 𝑢 + 2.8914 × 10−8 𝛽𝜌 𝑢 𝑢
𝑑𝑥 𝑘𝑎𝑔

 dp/dx = pressure gradient, psi/ft

 mg = gas viscosity, cp
 kag = permeability, md
 u = local fluid velocity, Mcuft/d·ft2
 b = Forchheimer coefficient, 1/ft
 r = gas density, lbm/cuft

43
Estimation of Forchheimer Coefficient

 The Forchheimer coefficient could be determined using several

techniques as listed below
 Analysis of well test data
 Experimental measurement in lab

 Empirical correlations

 We will review experimental methods and empirical correlations

44
High Velocity Liquid Flow Experiments

 Dp across linear core plugs is measured at increasing flow rates

 Forchheimer equation is manipulated to interpret the data set
 Rearranging Forchheimer equation as shown below
𝑑𝑝/𝑑𝑥 ∆𝑝/𝐿 1 𝜌𝑢
 − = = +𝛽
𝑢𝜇 𝑢𝜇 𝑘 𝜇
 Introducing two new flow groups
 𝐹𝐹𝐿 = 𝑑𝑝/𝑑𝑥 /𝑢𝜇 = ∆𝑝/𝑢𝜇𝐿
 𝑥𝐹𝐿 = 𝜌 𝑢 /𝜇

45
High Velocity Liquid Flow Experiments

1
 𝐹𝐹𝐿 = + 𝛽 𝑥𝐹𝐿
𝑘
 A Cartesian plot of FFL = Dp/umL vs xFL = r|u|/m
 We will call this plot as Forchheimer plot
 Test data should display a straight line
 The slope of the Forchheimer plot gives b
 The intercept of the Forchheimer plot is equal to 1/k
 𝑘 = 1/𝐹0𝐿
 𝛽 = 𝑚𝐹𝐿

46
High Velocity Liquid Flow in Porous Media- Example 8

 Reconsider the data given in Example 7

 m = 0.945 cp and r = 0.998 g/cc
 Construct the Forchheimer plot
 Predict the Forchheimer coefficient and permeability
 First, we compute the flow groups FFL = Dp/umL vs xFL = r|u|/m
 The calculated values are given in Table 7
 The unit for the flow group FFL = Dp/umL is atm·s/cp·cm2
 The unit for xFL = r|u|/m is g/(cp·cm2·s)

47
High Velocity Liquid Flow in Porous Media- Example 8
Table 7 – Forchheimer Plot Functions for Example 8
Test # |dp/dx| u m Dp/umL r|u|/m
() (atm/cm) (cm/s) (cp) (atm·s/cp·cm2) (g/cp·cm2·s)
72 0.00040918 3.41 0.955 0.00012547 3.5685
73 0.00052387 3.93 0.955 0.00013963 4.1053
74 0.00068041 4.61 0.955 0.00015459 4.8160
75 0.00080440 5.15 0.955 0.00016365 5.3783
76 0.00100124 5.99 0.956 0.00017475 6.2562
77 0.00121978 6.82 0.956 0.00018695 7.1245
78 0.00154681 7.69 0.953 0.00021117 8.0486
79 0.00175294 8.49 0.952 0.00021692 8.8981
80 0.00210322 9.49 0.955 0.00023203 9.9182
81 0.00254339 10.47 0.930 0.00026121 11.2348
82 0.00283633 11.20 0.931 0.00027192 12.0094
83 0.00331680 12.18 0.931 0.00029244 13.0583
84 0.00423899 13.85 0.931 0.00032885 14.8413
85 0.00514569 15.61 0.931 0.00035414 16.7293
86 0.00613763 17.37 0.931 0.00037957 18.6172

48
High Velocity Liquid Flow in Porous Media- Example 8

49
High Velocity Liquid Flow in Porous Media- Example 8

 Forchheimer plot is shown in previous figure

 The experimental data display a well defined straight line
 The intercept is F0L = 6.941x10-5 atm·s/cp·cm2
 The slope is mFL = 1.6958x10-5 atm·s2/g
1 1 cp∙cm2
 𝑘= = = 14,407 = 14,407 d
𝐹0𝐿 6.941×10−5 atm∙s/cp∙cm2 atm∙s
atm∙s2 1/cm 1
 𝛽 = 𝑚𝐹𝐿 = 1.6958 × 10−5 × 1.0134 × 106 = 17.19
g atm∙s2 /g cm

50
High Velocity Gas Flow Experiments

 The formulation of compressible gas flow under high-velocity flow

conditions are presented in Appendix A
 The final solution for Forchheimer plot for gas flow is
𝐴 𝑀𝑔 1 𝑤𝑠𝑐
 𝑝12 − 𝑝22 = +𝛽
2𝑤𝑠𝑐 𝜇𝑔 𝑧 𝑅 𝑇 𝐿 𝑘𝑎𝑔 𝐴 𝜇𝑔

 𝑤𝑠𝑐 = 𝜌𝑠𝑐 𝑞𝑠𝑐

 wsc = mass flow rate at standard conditions, g/s
 rsc = gas density at standard conditions, g/cm3
 qsc = volumetric flow rate at standard conditions, cm3/s

51
High Velocity Gas Flow Experiments

 Introduce two new flow groups

𝐴 𝑀𝑔
 𝐹𝐹𝑔 = 𝑝12 − 𝑝22
2 𝑤𝑠𝑐 𝜇𝑔 𝑧 𝑅 𝑇 𝐿

 𝑥𝐹𝑔 = 𝑤𝑠𝑐 /𝐴 𝜇𝑔
1
 𝐹𝐹𝑔 = + 𝛽 𝑥𝐹𝑔
𝑘𝑎𝑔

 a Cartesian plot of the flow groups FFg vs xFg

 Forchheimer plot for gas flow experiments

52
High Velocity Gas Flow Experiments

 F0g = intercept of Forchheimer plot for gas

 mFg = slope of Forchheimer plot for gas
 𝑘𝑎𝑔 = 1/𝐹0𝑔
 𝛽 = 𝑚𝐹𝑔

53
Empirical Correlations for Forchheimer Coefficient

 The literature is full of experimental and theoretical studies on the

nature of Forchheimer coefficient
 Correlate the Forchheimer coefficient to other physical rock properties
 Forchheimer coefficient correlates well with permeability, porosity, and
tortuosity
 Tortuosity is defined as the ratio of the average distance traveled by
fluid through porous rock to the core length
 Permeability, porosity, and Forchheimer coefficient measured on core
samples

54
Empirical Correlations for Forchheimer Coefficient

 Data set reported by Noman and Archer (1987)

 Their data set given in Table 8
 Construct cross plots of Forchheimer coefficient vs permeability and
Forchheimer coefficient vs porosity
 The plots are given in next two figures
 Forchheimer coefficient declines as permeability increase
 A well established linear trend on log-log scale
 A linear trend on the log-log scale suggest a power law relationship

55
Empirical Correlations for Forchheimer Coefficient
Table 8 – Data for Forchheimer coefficient Table 8 – Data for Forchheimer coefficient
(Adapted from Noman and Archer) (Adapted from Noman and Archer)
k f b k f b
(md) (fraction) (1/ft) (md) (fraction) (1/ft)
0.064 0.1231 4.246x1013 0.050 0.1231 1.021x1013
0.102 0.1150 1.740x1013 0.210 0.0934 3.709x1011
0.460 0.1001 8.650x1011 2.210 0.1217 1.730x1010
0.515 0.1100 7.260x1011 17.010 0.1692 2.021x109
33.000 0.2100 5.191x108 5.900 0.1490 7.379x109
15.200 0.2050 1.170x109 0.140 0.0587 7.931x1012
10.500 0.2090 2.010x109 0.230 0.0996 1.130x1012
13.000 0.2390 1.080x109 99.500 0.1517 7.660x108
32.500 0.2560 2.630x108 0.410 0.0743 8.028x1011
28.400 0.1960 6.361x108 0.041 0.0683 3.130x1013
20.800 0.1901 8.153x108 0.260 0.1139 1.140x1012
0.160 0.1005 1.190x1012 59.000 0.2076 3.679x108
0.170 0.1004 4.109x1011 2.100 0.1382 1.160x1010
0.170 0.0994 9.900x1011 0.520 0.1002 1.240x1011
0.110 0.1026 3.091x1012 0.038 0.0819 9.110x1013
0.140 0.1350 2.470x1012 0.700 0.1019 1.240x1011
0.050 0.1231 1.021x1013 4.030 0.0990 9.879x109

56
Empirical Correlations for Forchheimer Coefficient

57
Empirical Correlations for Forchheimer Coefficient

 𝛽 = 𝑎/𝑘 𝑏
 a and b are constant coefficients
 Forchheimer coefficient decreases while porosity rises
 A general linear trend on the log-log scale
 Forchheimer coefficient vs porosity relationship could be formulated as
 𝛽 = 𝑐/𝜙 𝑑
 Combining all the results
 𝛽 = 𝑎𝜏 𝑑 /𝑘 𝑏 𝜙 𝑐

58
Empirical Correlations for Forchheimer Coefficient

 Many empirical correlations

 Empirical correlations may be grouped as listed below
 Single-phase
 Gas flow at immobile liquid saturations

59
Empirical Correlations for Forchheimer Coefficient
Table 9 – Beta Correlations for Single Phase Flow

Investigators Equation Eq. #

2.2348 × 1010
Firoozabadi-Katz 𝛽= (78)
𝑘 1.201

6.15 × 1010
Jones 𝛽= (79)
𝑘 1.55

4.85 × 104
Geertsma 𝛽= (80)
𝑘 0.5 𝜙 5.5

5.5 × 109
Tek-Coats-Katz 𝛽 = 1.25 0.75 (81)
𝑘 𝜙

8.917 × 108
Liu-Civan-Evans 𝛽= 𝜏 (82)
𝑘𝜙

Cooper-Wang- 3.11 × 109 1.943

𝛽= 𝜏 (83)
Mohanty 𝑘1.023

4.115 × 108 3.35

Thauvin-Mohanty 𝛽= 𝜏 (84)
𝑘 0.98 𝜙 0.29

60
Empirical Correlations for Forchheimer Coefficient
Table 10 - Beta Correlations for Gas Flow at Immobile Liquid Saturation

Investigators Equation Eq. #

4.85 × 104
Geertsma 𝛽 = 0.5 5.5
(85)
𝑘𝑔 𝜙 1 − 𝑆𝐿

Noman- 1.22
𝜙 1 − 𝑆𝐿
Shrimanker- 𝛽 = 2.48 × 109 (86)
Archer 𝑘𝑔

Evans- 0.787
𝜇𝑔 / 𝜌𝑔 𝑝𝑜𝑣 − 𝑝𝑅 𝑘𝑔
Hudson- 𝛽 = 13.387 (87)
Greenlee 𝑘𝑔0.5 𝜙 1 − 𝑆𝐿

1.38 × 106
Kutasov 𝛽 = 0.5 1.5
(88)
𝑘𝑔 𝜙 1 − 𝑆𝐿

61
6.3 Variations of permeability
Variation of Permeability

 Permeability of porous rocks changes enormously

 For unconsolidated sandstone and carbonated rocks containing
microfractures and vugs, permeability may be higher than 1 d
 In case of consolidated tight sandstones, permeability values are on the
order of 0.1 md
 Shale formations exhibit extremely low permeability values in the order
of microdarcy (10-6 d) and nanodarcy(10-9 d)
 A general permeability classification of porous rocks in next table

63
Variation of Permeability

Table 12 – Formation classification based on

permeability values
Permeability Classification
(md) ()
> 1,000 Exceptional, rare
100-1,000 Very high
10-100 High
10-1 Normal
1-0.1 Low
0.1-0.01 Very low
<0.01 Extremely low

64
Variation of Permeability – Example 11

 Permeability varies substantially with the location throughout

subsurface reservoir rock
 As an example, we present the change in core plug permeability along
the vertical depth for Hiram Well # 17
 The measured core plug permeabilities are tabulated in Table 13
 Permeability data are plotted as a function of depth in next figure
 Notice that permeability axis is logarithmic

65
 Table 13 – Core plug permeability along depth, Hiram Well #17
Sample # Depth Permeability Sample # Depth Permeability
() (ft) (md) () (ft) (md)
1 2,880 1,271.0 25 2,904 49.0
2 2,881 1,239.0 26 2,905 36.0
3 2,882 1,184.0 27 2,906 23.0
4 2,883 1,891.0 28 2,907 20.0
5 2,884 1,500.0 29 2,908 0.1
6 2,885 1,271.0 30 2,909 56.0
7 2,886 1,565.0 31 2,910 49.0
8 2,887 1,325.0 32 2,911 26.0
9 2,888 967.0 33 2,912 33.0
10 2,889 717.0 34 2,913 26.0
11 2,890 728.0 35 2,914 36.0
12 2,891 554.0 36 2,915 42.0
13 2,892 130.0 37 2,916 33.0
14 2,893 218.0 38 2,917 39.0
15 2,894 466.0 39 2,918 52.0
16 2,895 684.0 40 2,919 56.0
17 2,896 600.0 41 2,920 33.0
18 2,897 336.0 42 2,921 46.0
19 2,898 150.0 43 2,922 36.0
20 2,899 277.0 44 2,923 29.0
21 2,900 78.0 45 2,924 33.0
22 2,901 101.0 46 2,925 23.0
23 2,902 82.0 47 2,926 33.0
24 2,903 82.0

66
Permeability Distribution Along Well Depth – Example 11

67
Variation of Permeability – Example 11

 Permeability varies significantly along the well depth

 The minimum permeability is 0.1 md
 The maximum permeability is 1,891 md
 The average permeability is 387.8 md
 The standard deviation of the permeability distribution is 527.5 md
 Four orders of magnitude difference between the maximum and
minimum permeability values
 Permeability is distributed log-normally

68
Variation of Permeability – Example 11

 Permeability generally decreases as the depth along the well increases

 At the top of the formation, in the interval from 2,880 ft to 2,890 ft,
logarithm of permeability values vary slightly
 In the interval from 2,890 ft to 2,905 ft, log permeability decreases
substantially, almost linearly
 In the interval from 2,905 ft to 2,926 ft, the log permeability along the
well depth changes slightly

69
Variation of Permeability – Example 11

 The change in the permeability may be due to the factors listed below
 Change in lithology
 Change in the grain size and grain size distribution

 Change in the clay content

 Cementation

 Change in porosity

 Sedimentation and historical geological events

70
Factors Controlling Permeability

 Many factors affect the permeability of porous rocks

 Porosity
 Microfractures and vugs

 Pore size distribution

 Lithology and mineral composition

 Burial depth

 Compaction and consolidation

 Confining pressure

 Cementation

 Clay content and clay swelling

71
Factors Controlling Permeability

 A general correlation between permeability and porosity

 Typically, permeability and porosity are directly proportional
 If porosity is large, permeability is also high
 Reservoir rocks are subject to large stresses
 Under high stresses, some reservoir rocks may be naturally fractured
 Some fractures may be several hundreds of feet long
 Some fractures may be several cm long
 Natural fractures in the order of 1 cm long are called microfractures

72
Factors Controlling Permeability

 Long fractures are treated differently than the microfractures

 Microfractures may enhance permeability of porous rocks substantially
 Some carbonated rocks contain vugs
 Vugs are very large pores in the order of 1 mm to 1 cm in diameter
 Especially, when vugs are interconnected and form a network,
permeability of porous rock may be improved considerably
 Permeability is proportional to square of the mean pore radius
 For well sorted grains, pore size is related to the grain size

73
Factors Controlling Permeability

 A porous rock with larger grains possesses higher permeability

 Smaller grain sizes lead to smaller pores and lower permeability
 In general, poorly sorted porous rocks have low permeability values
 Cementation reduces pore sizes and consequently lowers permeability
 Most reservoir rocks contains small amount of free clay particles
 Clay particles fill some of the pores
 Usually, internal surface area of pores is covered by clays
 Clays are sensitive to changes in salinity

74
Factors Controlling Permeability

 When pore filling clays are contacted by fresh water or aqueous fluids
with lesser salinity, clays swell and increase their volume
 Clay swelling may reduce permeability substantially
 In general, when burial depth increases, overburden stress acting on
reservoir rock increases
 Higher overburden stress leads to more consolidation of porous
material and reduces permeability
 As a general rule of thumb, permeability decreases with increasing
burial depth

75
Factors Controlling Permeability

 Physical properties of rocks are sensitive to net confining stress

 When net confining stress increases, both permeability and porosity
decrease
 Reduction on permeability as a function of net confining stress depends
on composition of porous rocks and degree of consolidation

76
Effect of Confining Pressure – Example 12

 Permeability of a consolidated sandstone core plug is measured at

several different net confining stress levels
 The measured permeabilities as a function of confining pressure are
given in Table 14
 Plot measured permeabilities as a function of confining pressure

 Permeability decreases as confining pressure increases

 Permeability varies nonlinearly with net confining stress

77
Effect of Confining Pressure – Example 12

Table 14 – Permeability as a
function of confining pressure
pc k
(psia) (md)
1,416 0.0370
1,945 0.0345
2,455 0.0331
2,900 0.0320
3,444 0.0319
3,927 0.0311
4,437 0.0307
4,925 0.0312

78
6.4 Permeability-porosity relationship
Permeability-Porosity Relationship

 Permeability and porosity of porous rocks are related

 In general, permeability is directly proportional to porosity
 High porosity values imply high permeability values
 Low permeability corresponds to low porosity

 Semi empirical models relating permeability and porosity

 Permeability-porosity cross plot

80
Semi-Empirical Models for Permeability-Porosity Relationship

 Some researchers have developed theoretical equations relating

permeability to porosity
 In these models, porous rocks are represented by a bundle of capillary
tubes
 There exist many semi empirical permeability models
 Some of the most common permeability models
 Kozeny-Carman
 Timur

 Wyllie and Rose

81
Semi-Empirical Models for Permeability-Porosity Relationship

 Kozeny-Carman model
 𝑘 = 𝑐𝑑𝑔2 𝜙 3 /72𝜏 1 − 𝜙 2

 c = a constant
 dg = mean grain diameter
 t = tortuosity
 Tortuosity is defined as the ratio of the average distance traveled by
fluid through porous rock to the core length

82
Semi-Empirical Models for Permeability-Porosity Relationship

 Timur presented the following semi empirical equation

 𝑘 = 0.136 𝜙 4.4 /𝑆𝑤𝑐2

 Swc = connate (irreducible) water saturation

 Wyllie and Rose model
 𝑘 = 100 𝜙 2 1 − 𝑆𝑤𝑐 /𝑆𝑤𝑐 2
 Semi empirical permeability models give some general trends
 Their applicability to real reservoir rocks is very limited

83
Permeability-Porosity Cross Plot
 For a given porosity value, permeability may vary widely
 For a specific reservoir, permeability and porosity may correlate well
 Observations on many sets of measured permeability and porosity data
 Relationship between permeability and porosity may be represented by
an equation of the form given below
 ln 𝑘 = 𝑦0 + 𝑚𝜙
 k in md, f in fraction
 The plot indicated by the eq. above is known as permeability-porosity
cross plot

84
Permeability-Porosity Cross Plot

 The expression ln 𝑘 = 𝑦0 + 𝑚𝜙 leads to a semilogarithmic plot

 Semilogarithmic plot should display an approximate straight line
 y0 and m are computed from the intercept and slope of straight line
 Permeability-porosity cross plot may also be alternatively expressed as
 𝑘 = 𝑧0 exp 𝑚𝜙
 𝑧0 = exp 𝑦0
 Exponential eq. is more convenient for fitting the measured data using
spreadsheet programs

85
Permeability-Porosity Cross Plot - Example 13

 Permeability and porosity values are measured on 47 core plugs from

Hiram Well #17
 Table 15 gives the measured data
 Construct the permeability-porosity cross plot
 Fit the measured data with an exponential function
 Obtain the coefficients of the fit

86
 Permeability-Porosity Cross Plot – Example 13
Table 15 - Permeability and porosity data from Hiram
Well #17
Sample
# Depth f k 31 2,910 0.173 49.0
() (ft) (fraction) (md) 32 2,911 0.178 26.0
1 2,880 0.289 1,271.0 16 2,895 0.249 684.0 33 2,912 0.178 33.0
2 2,881 0.285 1,239.0 17 2,896 0.272 600.0 34 2,913 0.156 26.0
3 2,882 0.281 1,184.0 18 2,897 0.237 336.0 35 2,914 0.174 36.0
4 2,883 0.288 1,891.0 19 2,898 0.219 150.0 36 2,915 0.173 42.0
5 2,884 0.279 1,500.0 20 2,899 0.220 277.0 37 2,916 0.168 33.0
6 2,885 0.292 1,271.0 21 2,900 0.194 78.0 38 2,917 0.166 39.0
7 2,886 0.290 1,565.0 22 2,901 0.174 101.0 39 2,918 0.172 52.0
8 2,887 0.297 1,325.0 23 2,902 0.184 82.0 40 2,919 0.169 56.0
9 2,888 0.274 967.0 24 2,903 0.167 82.0 41 2,920 0.152 33.0
10 2,889 0.278 717.0 25 2,904 0.169 49.0 42 2,921 0.161 46.0
11 2,890 0.280 728.0 26 2,905 0.171 36.0 43 2,922 0.174 36.0
12 2,891 0.222 554.0 27 2,906 0.159 23.0 44 2,923 0.148 29.0
13 2,892 0.203 130.0 28 2,907 0.165 20.0 45 2,924 0.157 33.0
14 2,893 0.215 218.0 29 2,908 0.130 0.1 46 2,925 0.157 23.0
15 2,894 0.255 466.0 30 2,909 0.168 56.0 47 2,926 0.156 33.0

87
Permeability-Porosity Cross Plot – Example 13

88
Permeability-Porosity Cross Plot – Example 13

 Log(k) versus f displays a well defined straight line

 Using a spreadsheet program, an exponential equation is fit to the
permeability-porosity data
 The exponential equation has the following form
 𝑘 = 0.1584 exp 31.954𝜙
 𝑚 = 31.954
 𝑧0 = 0.1584
 𝑦0 = ln 𝑧0 = ln 0.1584 = −1.84

89
References

 American Petroleum Institute, Recommended Practice RP 40,

Recommendation Practices for Core Analysis, Second Edition, API
Publishing Services, Washington DC, 1998
 Amyx, J.W. et al.: Petroleum Reservoir Engineering, McGraw-Hill, Inc.,
1960
 Ahmed, T.: Reservoir Engineering Handbook, Fourth Edition, Gulf
Professional Publishing, 2010
 Dandekar, A.Y.: Petroleum Reservoir Rock and Fluid Properties, Taylor
and Francis Group, LLC, Boca Raton, 2006

90
References
 McPhee, C., Reed, J., and Zubizarreta, I.: Core Analysis: A Best Practice
Guide, Elsevier, Amsterdam, 2015
 Monicard, R.P.: Properties of Reservoir Rocks: Core Analysis, Gulf
Publishing Company, Houston, 1980
 Peters, E.J: Advanced Petrophysics: Volume 1: Geology, Porosity,
Absolute Permeability, Heterogeneity, and Geostatistics, Live Oak Book
Company, Austin, Texas, 2012
 Tiab, D. and Donaldson, E.C.: Petrophysics, Second Edition, Gulf
Professional Publishing, 2004

91
Useful Links

 http://petrowiki.org/Permeability_determination
 http://petrowiki.org/Permeability_estimation_in_tight_gas_reservoirs
 http://petrowiki.org/Core_analyses_in_tight_gas_reservoirs
 http://petrowiki.org/Statistical_data_correlations_in_tight_gas_reservo
irs
 https://en.wikipedia.org/wiki/Permeability_(earth_sciences)
 https://en.wikipedia.org/wiki/Klinkenberg_correction
 http://perminc.com/resources/fundamentals-of-fluid-flow-in-porous-
media/chapter-2-the-porous-medium/permeability/

92
Useful Links

 https://www.youtube.com/watch?v=hgZHaqYgmRg
 https://www.youtube.com/watch?v=nV5VJEdryJU
 https://www.youtube.com/watch?v=Q4LBnWg-87E
 https://www.youtube.com/watch?v=Eur_qpTKzrA
 https://www.youtube.com/watch?v=v5ItNoZz42c
 https://www.youtube.com/watch?v=9NLZhftPE9M
 https://www.youtube.com/watch?v=bsgTBHuXiSk
 https://www.youtube.com/watch?v=8mfBomrw0rs

93
Appendix A – Formulation of High-
Velocity Gas Flow in a Linear Core
Formulation of High-Velocity Gas Flow in a Linear Core

 Forchheimer equation for steady state flow

𝑑𝑝 𝜇𝑔
 − = 𝑢 + 𝛽𝜌 𝑢 𝑢
𝑑𝑥 𝑘𝑎𝑔

 𝑢 = 𝑞/𝐴
 Combining the eqs. Above
𝑑𝑝 𝜇𝑔 𝑞 𝑞 𝑞
 − = + 𝛽𝜌
𝑑𝑥 𝑘𝑎𝑔 𝐴 𝐴 𝐴

 Multiplying eq. above by r

𝑑𝑝 𝜇𝑔 𝜌𝑞 𝜌𝑞 𝜌𝑞
 −𝜌 = +𝛽
𝑑𝑥 𝑘𝑎𝑔 𝐴 𝐴 𝐴

95
Formulation of High-Velocity Gas Flow in a Linear Core

 Mass rate is constant and equal to mass flow rate at standard cond.
 𝑤 = 𝜌𝑞 = 𝑤𝑠𝑐 = 𝜌𝑠𝑐 𝑞𝑠𝑐
 Substituting the last eq. into Forchheimer eq.
𝑑𝑝 𝜇𝑔 𝑤𝑠𝑐 𝑤𝑠𝑐 𝑤𝑠𝑐 𝑤𝑠𝑐 𝜇𝑔 1 𝑤𝑠𝑐
 −𝜌 = +𝛽 = +𝛽
𝑑𝑥 𝑘𝑎𝑔 𝐴 𝐴 𝐴 𝐴 𝑘𝑎𝑔 𝐴 𝜇𝑔

 Using real gas EOS

 𝜌 = 𝑝𝑀𝑔 /𝑧𝑅𝑇
 Substituting in high velocity flow eq.

96
Formulation of High-Velocity Gas Flow in a Linear Core

𝑝𝑀𝑔 𝑑𝑝 𝑤𝑠𝑐 𝜇𝑔 1 𝑤𝑠𝑐

 − = +𝛽
𝑧𝑅𝑇 𝑑𝑥 𝐴 𝑘𝑎𝑔 𝐴 𝜇𝑔

 Rearranging
𝐴 𝑀𝑔 𝑝 1 𝑤𝑠𝑐
 − 𝑑𝑝 = +𝛽 𝑑𝑥
𝑤𝑠𝑐 𝑅 𝑇 𝜇𝑔 𝑧 𝑘𝑎𝑔 𝐴 𝜇𝑔

 Considering isothermal flow and integrating

𝐴 𝑀𝑔 𝑝2 𝑝 𝐿 1 𝑤𝑠𝑐
 − 𝑝1 𝜇𝑔 𝑧
𝑑𝑝 = 0
+𝛽 𝑑𝑥
𝑤𝑠𝑐 𝑅 𝑇 𝑘𝑎𝑔 𝐴 𝜇𝑔

 𝜇𝑔 𝑧 ≈ constant = 𝜇𝑔 𝑧

97
Formulation of High-Velocity Gas Flow in a Linear Core

𝐴 𝑀𝑔 𝑝2 1 𝑤𝑠𝑐 𝐿
 − 𝑝1
𝑝𝑑𝑝 = +𝛽 0
𝑑𝑥
𝑤𝑠𝑐 𝜇𝑔 𝑧 𝑅 𝑇 𝑘𝑎𝑔 𝐴 𝜇𝑔

 Evaluating the integrals

𝐴 𝑀𝑔 𝑝2 𝑝2 1 𝑤𝑠𝑐
 − |𝑝1 = +𝛽 𝑥|𝐿0
𝑤𝑠𝑐 𝜇𝑔 𝑧 𝑅 𝑇 2 𝑘𝑎𝑔 𝐴 𝜇𝑔
𝐴 𝑀𝑔 1 𝑤𝑠𝑐
 − 𝑝12 − 𝑝22 = +𝛽 𝐿
2 𝑤𝑠𝑐 𝜇𝑔 𝑧 𝑅 𝑇 𝑘𝑎𝑔 𝐴 𝜇𝑔

 Rearranging
𝐴 𝑀𝑔 1 𝑤𝑠𝑐
 − 𝑝12 − 𝑝22 = +𝛽
2 𝑤𝑠𝑐 𝜇𝑔 𝑧 𝑅 𝑇 𝐿 𝑘𝑎𝑔 𝐴 𝜇𝑔

98