1.

Dimensionless Group Analysis

Table of Content
1. Dimensionless Group Analysis ..................................................................................................................... 1 1.1. Introduction....................................................................................................................................................................................... 4 1.2. Methods for the Derivation of Dimensionless Groups ................................................................................ 4 1.1.1. Dimensional Analysis ................................................................................................................................................... 4 1.2.1.1. The Buckingham's Pi-Theorem ................................................................................................................. 5 1.2.2. Inspectional Analysis..................................................................................................................................................... 7 1.3. Conventional Dimensionless groups ........................................................................................................................... 8 1.3.1. Ratio of Convection and Diffusion .................................................................................................................... 8 1.3.2. Dimensionless Groups Involving Surface Tension........................................................................... 13 1.3.3. Dimensionless Groups Involving Gravity ................................................................................................ 14 1.3.4. Dimensionless Numbers in Natural Convection ................................................................................. 15 1.3.5. Other Dimensionless Groups ............................................................................................................................... 16 1.4. Dimensionless Groups in Petroleum Engineering ........................................................................................ 19 1.4.1. Dombrowski-Brownell number ......................................................................................................................... 19 1.4.2. Stability number .............................................................................................................................................................. 20 1.4.3. Lake number....................................................................................................................................................................... 21 1.4.4. Capillary end-effect number ................................................................................................................................. 22 1.4.5. Capillary-to-gravity number ................................................................................................................................. 22 1.4.6. Gravity number ................................................................................................................................................................ 23 1.5. Dimensionless Groups for Fluid Flow in Fractured Porous Medium .......................................... 23 1.5.1. Fracture Capillary Number .................................................................................................................................... 23 1.5.2. Modified Diffusive Capillary Number ........................................................................................................ 24 1.5.3. Fracture Diffusion Index ......................................................................................................................................... 24 1.5.4. Matrix-Fracture Diffusion Number ................................................................................................................ 25 1.5.5. Dimensionless Time .................................................................................................................................................... 26 1.6. New Dimensionless Groups as Combination of Dimensionless Numbers .............................. 29 1.6.1. Grattoni et al. Number (2001) ............................................................................................................................. 29 1.6.2. Kulkarni and Rao Number (2006) ................................................................................................................... 29 1.6.3. Rostami et al. Number (2009) ............................................................................................................................. 30 1.6.4. Rostami et al. Number (2010) ............................................................................................................................. 30 1.7. Dimensionless Groups for Scaling EOR Processes ..................................................................................... 31 1.8. Application ..................................................................................................................................................................................... 35 2

1.9. References ....................................................................................................................................................................................... 40

1.1. Introduction
Dimensionless groups are products of suitable powers of important variables which have zero net dimension. They are widely used in mathematics, physics and engineering. The dimensionless groups are useful for several reasons. They reduce the number of variables needed for description of the problem. They can thus be used for reducing the amount of experimental data and at making correlations. They predicting the effect of changing one of the individual variables in a process (which it may be impossible to vary much in available equipment) by determining the effect of varying the dimensionless group containing this parameter (this must be done with some caution, however). They make the results independent of the scale of the system and of the system of units being used. They simplify the scaling-up or scaling-down of results obtained with models of systems by generalizing the conditions which must exist for similarity between a system and its model. They deduce variation in importance of mechanisms in a process from the numerical values of the dimensionless groups involved; for instance, an increase in the Reynolds number in a flow process indicates that molecular (viscous) transfer mechanisms will be less important relative to transfer by bulk flow (inertia effects), since the Reynolds number is known to represent a measure of the ratio of inertia forces to viscous forces.

1.2. Methods for the Derivation of Dimensionless Groups

Dimensionless groups can be attained in two ways: dimensional analysis and inspectional analysis. The dimensional analysis approach is based on the Buckingham's Pi-theorem. Inspectional analysis is better since it takes advantage of the problems' full mathematical specification based on physical laws, and reveals a higher degree of similarity than dimensional analysis.

1.1.1. Dimensional Analysis

Dimensional analysis is a mathematical technique making use of the study of dimensions. In dimensional analysis, from a general understanding of fluid phenomena, one first predicts the physical parameters that will influence the flow, and then, by grouping these parameters in dimensionless combinations, a better understanding of the flow phenomena is made possible.

First step in dimensional analysis is to identify the transport properties (fluid velocity, heat influx, mass influx, etc.) and the material properties (viscosity, thermal conductivity, specific heat, density, etc.) which are relevant to the problem, and list out all these quantities along with their dimensions. In general, if a problem contains n important variables, and there groups contain m fundamental dimensions, then there are n-m dimensionless groups. This important principle is called the Buckingham Pi Theorem. It should be emphasized that dimensional analysis does not provide a complete solution to fluid problems. It provides a partial solution only. The success of dimensional analysis depends entirely on the ability of the individual using it to define the parameters that are applicable. If one omits an important variable, the results are incomplete and this may lead to incorrect conclusions. On the other hand, if one includes a variable that is totally unrelated to the problem, an additional insignificant dimensionless group will result.

1.2.1.1. The Buckingham's Pi-Theorem There are several methods of reducing a number of dimensional variables into a smaller number of dimensionless groups. The scheme given here was proposed in 1914 by Buckingham and is now called the Buckingham pi theorem. The name pi comes from the mathematical notation , meaning a product of variables. The dimensionless groups found from the theorem are power products denoted by 1, 2, 3, etc. The method allows the pi's to be found in sequential order without resorting to free exponents. The first part of the pi theorem explains what reduction in variables to expect: If a physical process satisfies the PDH1 and involves n dimensional variables, it can be reduced to a relation between only k dimensionless variables or s. The reduction j= n - k equals the maximum number of variables which do not form a pi among themselves and is always less than or equal to the number of dimensions describing the variables. Take the specific case of force on an immersed body contains five variables F, L, U, and described by three dimensions {MLT}. Thus n = 5 and j 3. Therefore it is a good guess that we can reduce the problem to k pi's, with k = n - j 5 -3 = 2. On rare occasions it may take more pi's than this minimum.

Principle of Dimensional Homogeneity: If an equation truly expresses a proper relationship between variables in a physical process, it will be dimensionally homogeneous; i.e., each of its additive terms will have the same dimensions.

The second part of the theorem shows how to find the pi's one at a time: Find the reduction j, then select j scaling variables which do not form a pi among themselves. Each desired pi group will be a power product of these j variables plus one additional variable which is assigned any convenient nonzero exponent. Each pi group thus found is independent. To be specific, suppose that the process involves five variables
1 = f (2, 3, 4, 5) (1-1)

Suppose that there are three dimensions {MLT} and we search around and find that indeed j = 3. Then k = 5- 3= 2 and we expect, from the theorem, two and only two pi groups. Pick out three convenient variables which do not form a pi, and suppose these turn out to be 2, 3 and 4. Then the two pi groups are formed by power products of these three plus one additional variable, either 1 or 5:
1= (2)a(3)b(4)c 1=M0L0T0 2= (2)a(3)b(4)c 5=M0L0T0 (1-2)

Here we have arbitrarily chosen 1 and 5, the added variables, to have unit exponents. Equating exponents of the various dimensions is guaranteed by the theorem to give unique values of a, b, and c for each pi. And they are independent because only 1 contains 1 and only 2 contains 5. It is a very neat system once you get used to the procedure. Typically, six steps are involved: 1. List and count the n variables involved in the problem. If any important variables are missing, dimensional analysis will fail. 2. List the dimensions of each variable according to {MLT} or {FLT}. 3. Find j. Initially guess j equal to the number of different dimensions present, and look for j variables which do not form a pi product. If no luck, reduce j by 1and look again. With practice, you will find j rapidly. 4. Select j scaling parameters which do not form a pi product. Make sure they please you and have some generality if possible, because they will then appear in every one of your pi groups. Pick density or velocity or length. Do not pick surface tension, e.g., or you will form six different independent Weber-number parameters and thoroughly annoy your colleagues. 5. Add one additional variable to your j repeating variables, and form a power product. Algebraically find the exponents which make the product dimensionless. Try to arrange for 6

your output or dependent variables (force, pressure drop, torque, power) to appear in the numerator, and your plots will look better. Do this sequentially, adding one new variable each time, and you will find all n - j = k desired pi products. 6. Write the final dimensionless function, and check your work to make sure all pi groups are dimensionless.

1.2.2. Inspectional Analysis

Inspectional analysis is a similar method for obtaining dimensionless groups to study the mechanistic behavior of a process. However, IA transforms the dimensional space to dimensionless space variable-by-variable instead of making decision based on the primary dimensions of variables (Ruark, 1935; Bear, 1972; Shook, 1992; Djuro, 2002). Inspectional analysis combines differential equations of the physical process together with its initial and boundary conditions and then transforms the whole system into dimensionless forms. This can be done in the following steps: 1. Formulation of all governing equations together with the initial and boundary conditions 2. Transformation to dimensionless space 3. Primary elimination and secondary substitution 4. Redundancy elimination 5. Independence testing Both methods have their own advantages and limitations. Inspectional analysis usually gives scaling groups whose physical meanings are readily apparent, whereas the physical meanings of groups from dimensional analysis may be quite obscure (Ruark, 1935; Bear, 1972; Sonin, 1997; Gharbi et al, 1998; Djuro, 2002; Hernandez and Wojtanowicz, 2007). For example, without knowing a differential equation, it is easy to define intuitively a similarity group as a ratio of the capillary to viscous forces, or the viscous to gravity forces. On the other hand, IA requires mathematical equations for the process under study. If such equations are unavailable, inspectional analysis cannot begin. Dimensional analysis is useful in providing some guidance for setting up experiments to initiate the study and it is especially useful when people dont know much about their research objectives. Therefore, the dimensional analysis method is more universal than the

inspectional analysis when very little theory is available, dimensional analysis can be relied upon to provide initial guidance in setting up experiments. The inspectional analysis method is used more frequently in reservoir engineering because the underlying physical laws are known and expressed in the form of partial differential equations and boundary conditions (Shook et al, 1992). Principles of scaling in petroleum studies were outlined by Rapoport (1955) for the immiscible displacement of oil by water. Interestingly, different dimensionless groups have been derived by various authors, to describe the same phenomenon (Croes and Schwarz, 1955; Rapoport, 1955). In many cases these dimensionless groups were reportedly obtained from IA, but the procedure used was not explained.

1.3. Conventional Dimensionless groups

Dimensionless groups can be classified into different types.

1.3.1. Ratio of Convection and Diffusion

Diffusion is a molecular process, where material is transported down concentration gradients due to molecular motion. The molecules in a fluid are in constant motion which is random with equal probability in all directions. Consequently, at equilibrium, there is no net motion of mass in any one direction. However, if there is a spatial variation of the concentration of a species or the thermal energy of the molecules, this random motion will transport mass or heat down the gradient. Consequently, diffusive transport of mass, heat and momentum have some common characteristics. The diffusive transport (per unit time) of a quantity such as mass or heat through a material is directly proportional to the cross sectional area, inversely proportional to the length and directly proportional to the difference in concentration or temperature across the material. 1. For transport of mass, the mass flux is related to the concentration difference by Fick's law. If a concentration difference c is maintained between two ends of a slab of length l, the mass flux jc per unit area, which has units of (mass / area / time), is
=

(1-3)

Here, just from dimensional analysis, it is easy to see that the diffusion coefficient has units of L2T-1. 2. For the transport of heat, the heat flux is related to the temperature difference by Fick's law. If the temperature difference between two surfaces is T, then the heat flux je (energy/ area / time) is given by
=

(1-4)

where k is the thermal conductivity. It is possible to define a diffusion coefficient for heat transfer in the following form. Consider a volume which is heated up by a flux incident on it from its boundaries. The difference in temperature T can be expressed in terms of the difference in the specific energy between the two sides as T = (E/cv) where E is the specific energy (per unit volume). With this, the equation for the heat flux can be written as
=

(1-5)

It is obvious that the above equation has the same form as the mass flux equation, with a thermal diffusivity Dt = (k/cp) which has units of L2T-1. 3. The relation for the momentum diffusivity is a little more complicated. It turns out that the diffusion of momentum is due to viscous effects, and the diffusivity for momentum transfer is the "kinematic viscosity" = (/), which has units of L2T-1. Convective transport takes place due to the mean flow of a fluid. For example, if the flow has a velocity U and the difference in concentration between two points is c, the rate of transport of material per unit area is equal to the Uc. Consequently, the ratio between the rate of transport due to convective and diffusive effects is (Ul/D), where l is the length scale in the problem. Many of the dimensionless numbers provide ratios of convective to diffusive transport, or ratio of diffusive transport of two different quantities, as shown in Table 1.1.

Table 1.1: Dimensionless numbers as ratios of forces Dimensionless number Reynolds number Prandtl Number Schmidt number Peclet Number Momentum convection Momentum difussion Momentum difussion Heat diffusion Momentum difussion Mass diffusion Heat convection Heat diusion Ratio Expression = =

Reynolds number: It is named after Osborne Reynolds (18421912), a British engineer who first proposed it in 1883. The Reynolds number is always important, with or without a free surface, and can be neglected only in flow regions away from high-velocity gradients, e.g., away from solid surfaces, jets, or wakes. The Reynolds number is the ratio of the inertial to the viscous forces:
= Momentum convection = Momentum difussion

(1-6)

Re << 1, the viscous forces (friction) are dominant. The convective term can be neglected compared to the viscous term. Re >> 1, convection (or inertial) forces are dominant. But one should be really cautious here: in this case, the viscous forces cannot be neglected everywhere in the flow, only away from interfaces. In particular, in areas close to interfaces and boundary layers, the viscous forces are of the same order as the inertial forces, because in the vicinity of interfaces, very strong gradients may exist. The critical Reynolds number denotes the boundary between laminar and turbulent flow (it is 2000 for flow down a pipe). Low Re indicate laminar flow ("creeping" flows typically have Re < 1). Large Reynolds numbers indicate turbulent flow. The Modified Reynolds number is used to determine the flow regime of the fluid within the porous medium. Modification to the fluid velocity term (v) and the characteristic linear dimension (D) are required. When considering flow within the bed the appropriate velocity is the interstitial, which can be related to the superficial velocity V by (v = V/). The characteristic linear dimension was deduced by Kozeny and is the volume open to the fluid

10

flow divided by the surface area over which it must flow (i.e. product of volume of solids and specific surface area per unit volume)
= = 1 1

(1-7)

Thus, the Modified Reynolds number (Rem) can be expressed as:

= = 1 1

(1-8)

Conceptually, the number still represents the ratio of inertial to viscous forces in the fluid and provides a means to assess when the inertial effects become significant. The conventionally applied threshold to indicate significant turbulence is 2, whereas for the flow Reynolds number the conventional threshold is about 2000. It is important to note that the density term in above equation is the density of the fluid: the turbulences described are that of the fluid, the particles do not move in a packed bed. Prandtl Number: It is named after Ludwig Prandtl, who introduced the concept of boundary layer in 1904 and made significant contributions to boundary layer theory. Prandtl number is the ratio of momentum diffusivity (kinematic viscosity) to thermal diffusivity.
= Momentum difussion = = Heat diffusion

(1-9)

The Prandtl numbers of fluids range from less than 0.01 for liquid metals to more than 100,000 for heavy oils (Table 1.2). Note that the Prandtl number is in the order of 10 for water. The Prandtl numbers of gases are about 1, which indicates that both momentum and heat dissipate through the fluid at about the same rate. Heat diffuses very quickly in liquid metals (Pr 1) and very slowly in oils (Pr 1) relative to momentum. Consequently the thermal boundary layer is much thicker for liquid metals and much thinner for oils relative to the velocity boundary layer.

Table 1.2: Typical ranges of Prandtl numbers for common fluids.

11

Fluid Liquid metals Gases Water Light organic fluids Oils Glycerin

Pr 0.004-0.03 0.7-1.0 1.7-13.7 5-50 50-100000 2000-100000

Schmidt number: It was named after the German engineer Ernst Heinrich Wilhelm Schmidt (1892-1975). Schmidt number is a dimensionless number defined as the ratio of momentum diffusivity (viscosity) and mass diffusivity, and is used to characterize fluid flows in which there are simultaneous momentum and mass diffusion convection processes.
= Momentum diusion = = Mass diffusion

(1-10)

Where is the kinematic viscosity or (/), D is the mass diffusivity, is the dynamic viscosity of the fluid and is the density of the fluid. It physically relates the relative thickness of the hydrodynamic layer and mass-transfer boundary layer. The heat transfer analog of the Schmidt number is the Prandtl number. Peclet and PecletII Numbers The Peclet and PecletII numbers are analogous to the Reynolds number but for heat and mass transfer, respectively. The Peclet number is the ratio,
= Heat convection = Heat diusion

(1-11)

Using the Prandtl number, we can write the Peclet number as a function of Reynolds number:
= . (1-12)

Likewise, for mass transfer,

12

Mass convection Mass diusion

(1-13)

1.3.2. Dimensionless Groups Involving Surface Tension

There are three important dimensionless groups involving surface tension, which are the Weber number, the Capillary number and Bond number. Weber number: It is named after Moritz Weber (18711951) of the Polytechnic Institute of Berlin, who developed the laws of similitude in their modern form. The Weber number is the ratio of inertial and surface tension forces,
=

(1-14)

The Weber number is important only if it is of order unity or less, which typically occurs when the surface curvature is comparable in size to the liquid depth, e.g., in droplets, capillary flows, ripple waves, and very small hydraulic models. If We is large, its effect may be neglected. Capillary number: It is defined as the ratio of the viscous force to the capillary force at the pore scale (Chatzis and Morrow, 1984). For two-phases, it can be defined as:
= =

(1-15)

where v is the Darcy velocity, is the viscosity of the displacing phase, is the interfacial tension, k is the medium permeability, P is the pressure difference and L is the characteristic lenght . Morrow et al. 1988 have shown that variations in the pore size distribution or the degree of consolidation, produce changes in the value of the capillary number due to change of trapped oil size distribution. The influence of the pore structure and wettability could be introduced by using the capillary pressure term defined as:
= 2 cos

(1-16)

13

where is the contact angle and Ra is the average pore throat radius. For strongly wetting conditions, cos =1. At constant velocity, the capillary number is naturally constant, but when the flow rate varies with time it becomes a dynamic parameter. Capillary number affects the residual saturation of non-wetting phase in two phase flow for Nc >10-5 (Abrams, 1975). Lefebvre du Prey (1973) have showed that the capillary number affects the wetting phase relative permeability and residual saturation for Nc >10-7. Bond number: The Bond number (Hove et al., 1995) is a dimensionless group that measures the relative strength of gravity or buoyancy in some cases and capillary forces. For the trapping of the non-wetting phase, it can be expressed as
=

(1-17)

where is the density difference between the two fluids, g is the gravitational constant, l is a characteristic length of the porous medium often taken as the average grain radius, (Morrow et al., 1988). For a vertical flooding, NB takes into account the balance between gravity and capillary forces and is directly proportional to the advance of the displacing phase front. Thus, NB can also be a dynamic parameter depending additionally on the front velocity.

1.3.3. Dimensionless Groups Involving Gravity

Froude number: It is named after William Froude (18101879), a British naval architect who, with his son Robert, developed the ship-model towing-tank concept and proposed similarity rules for free-surface flows (ship resistance, surface waves, and open channels). Froude number is the ratio of inertial and gravitational forces:
=

(1-18)

where U is the velocity, and g is the acceleration due to gravity, and L is the characteristic length. Fr >> 1, the body forces can be neglected.

14

 Fr << 1, the body forces are significant. This number is important for flows with a free surface, such as in the dynamics between the waves and a boat. In applications involving rotation of fluids, the Froude number is also the ratio of centrifugal and gravitational forces. The Froude number is the dominant effect in free-surface flows and is totally unimportant if there is no free surface.

1.3.4. Dimensionless Numbers in Natural Convection

In natural convection, the driving force for convection is the body force caused by a variation in the density of the fluid, which is in turn caused by variation in temperature. In contrast to natural convection, the characteristic flow velocity is not known a priori, but has to be determined from the driving force due to the density difference. If we assume there is a balance between the driving force and the viscous stresses in the fluid, dimensional analysis can be used to infer that the characteristic velocity should scale as Uc (fD2/), where f is the driving force for convection per unit volume of fluid, D is the diameter of the body or the characteristic length, and is the viscosity. The driving force for convection, per unit volume of the fluid, is proportional to the product of the density variation and the acceleration due to gravity g. The density variation is proportional to T, where T is the variation in temperature and is the coefficient of thermal expansion. Therefore, the force per unit area is proportional to gT. With this, the characteristic velocity for the flow is Uc (D2gT/). The "Grashof number" is the Reynolds number based upon this convection velocity,
= =

(1-19)

In defining the Grashof number, we used the momentum diffusivity and the body force to obtain the characteristic velocity scale. An alternative definition is the Rayleigh number, where the thermal diffusivity (k/Cp) is used to determine the characteristic velocity. The characteristic velocity is then given by Uc (fD2Cp/k), where f =T is the driving force for convection per unit volume of fluid. The Reynolds number based on this convection velocity and momentum diffusivity is the Rayleigh number,

15

(1-20)

The "Rayleigh-Benard" instability of a fluid layer heated from below occurs when the Rayleigh number increases beyond a critical value.

1.3.5. Other Dimensionless Groups

Mach number: The Mach number is named after Ernst Mach (18381916), an Austrian physicist. The ratio is of fluid speed to speed of sound is relevant in two areas:
=

(1-21)

It is the most important correlating parameter when fluid velocities are near the local sonic velocity. Compressible fluids can be treated like incompressible fluids when Ma < 0.3. Euler number (pressure coefficient): This is named after Leonhard Euler (17071783) and is rarely important unless the pressure drops low enough to cause vapor formation (cavitation) in a liquid. The Euler number comes from the ratio between pressure forces and inertial forces,
=

(1-22)

Knudsen Number: If a flow is sufficiently rarefied, a dimensionless ratio called the Knudsen number is likely to come into prominence. Knudsen number is normally defined as the ratio between the mean free path of the molecules and a characteristic length. When the mean free path is very small compared with other lengths involved, as is the case with ordinary densities and sizes, Knudsen number does not enter; but when the mean free path becomes sufficiently large, the equations we have used to describe the flow are no longer entirely appropriate. Knudsen diffusion occurs when gas molecules experience significant interaction with pore walls. Knudsen diffusion is significant in porous media with very small pores (on the order of a few micrometers or smaller) or at low pressures. It is characterized by the ratio of the molecules mean free path divided by the average pore throat radius, 16

(1-23)

where Kn is the Knudsen number, is the gas phase molecular mean free path in m, and rpore is the characteristic length scale of the flow path in m, in porous media representing an the flow is non-Darcy. For Knudsen numbers less than 0.01 imply Darcy flow. For Knudsen numbers higher than 0.01, the flow equations must be modified for Knudsen diffusion and/or slip flow. Nusselt Number: It is named after Wilhelm Nusselt, who made significant contributions to convective heat transfer in the first half of the twentieth century, and it is viewed as the dimensionless convection heat transfer coefficient. It, therefore, provides a measure of the convection heat transfer at the surface. It is defined as:
=

equivalent hydraulic radius (Civan 2008). The Knudsen number signifies the degree to which

(1-24)

where, h is the heat transfer coefficient, Lc is a characteristic length and k is the thermal conductivity. To understand the physical significance of the Nusselt number, consider a fluid layer of thickness L and temperature difference T=T2-T1. Heat transfer through the fluid layer will be by convection when the fluid involves some motion and by conduction when the fluid layer is motionless. Heat flux (the rate of heat transfer per unit time per unit surface area) in either case will be
. = (1-25)

and
. =

(1-26)

Taking their ratio gives 17

. = = .

(1-27)

which is the Nusselt number. Therefore, the Nusselt number represents the enhancement of heat transfer through a fluid layer as a result of convection relative to conduction across the same fluid layer. The larger the Nusselt number, the more effective the convection. A Nusselt number of Nu=1 for a fluid layer represents heat transfer across the layer by pure conduction. Lewis number: It is named after Warren K. Lewis (18821975). Lewis number is a dimensionless number defined as the ratio of thermal diffusivity to mass diffusivity. It is used to characterize fluid flows where there is simultaneous heat and mass transfer by convection.
= =

(1-28)

where is the thermal diffusivity and D is the mass diffusivity. The Lewis number can also be expressed in terms of the Schmidt number and the Prandtl number:
= Sc Pr

(1-29)

Sherwood number: It is named in honor of T. K. Sherwood. Sherwood number represents the ratio between mass transfer by convection and mass transfer by diffusion.
= Convective mass transport = Diffusitive mass transport

(1-30)

where h is the mass transfer coefficient, Lc is the characteristic length, and D is the diffusion coefficient. In mass transfer, the Sherwood number plays the role the Nusselt number plays in heat transfer.

18

Stanton number: It is named after Thomas Edward Stanton. The Stanton number is the ratio of heat transferred into a fluid to the thermal capacity of fluid.
= Actual heat lux to the luid = Heat lux capacity of the luid low

(1-31)

where h is the convection heat transfer coefficient, is the density of the fluid, Cp is the specific heat of the fluid and v is the velocity of the fluid. Biot number: It is named after J. B. Biot who, in 1804, analyzed the interaction between conduction in a solid and convection at its surface. Biot number is the ratio of the convection at the surface to conduction within the body.
= Convection at the surface of the body = Conduction within the body

(1-32)

where h is the convective heat transfer coefficient, LC is the characteristic length and k is the thermal conductivity of the body. A small value of Bi indicates that the inner resistance of the body to heat conduction is small relative to the resistance to convection between the surface and the fluid. As a result, the temperature distribution within the solid becomes fairly uniform.

1.4. Dimensionless Groups in Petroleum Engineering

1.4.1. Dombrowski-Brownell number
In 1954, Dombrowski and Brownell presented a general correlation between residual wettingphase saturation and the dimensionless viscous and gravitational forces that cause desaturation. The dimensionless number they presented generally called the DombrowskiBrownell number, NDB. The DombrowskiBrownell number is defined as the ratio of the gravity force to the capillary force at the pore scale.
=

(1-33)

19

where is the fluid density difference, g is the effective gravitational constant, K is the medium permeability and is the interfacial tension. Dombrowski and Brownell presented wetting-phase saturation data that suggested a critical dimensionless number of 10-3. Above that number, the residual wetting-phase saturation was significantly reduced and approached zero. In two-phase flow, this number affects the residual saturation of non- wetting phase for NDB>10-5 (Morrow and Chatzis, 1984) and residual saturation of wetting phase for NDB>10-2. NDB can be interpreted as a microscopic version of the Bond number. Typical values for macroscopic Bond number suggest that no correction is needed in residual oil saturation in order to scale between lab and field condition. However, typical values for microscopic Bond number, NDB, suggest that capillary forces cannot be neglected on the scale of fluid flow. This is true primarily when saturation variation occurs in the sample, although this shows up nowhere in the definition of NDB. In fact, NDB can be used as an indicator when saturation variation expected. When present, it must be accounted for in the interpretation of relative permeability, especially in the range of low saturation (Edward, 1998).

1.4.2. Stability number

Peters and Flock (1981) developed a dimensionless stability number for quantitative prediction of the onset of instability in a displacement process. Their stability number for a cylindrical system is given by:
= 1 cos 1

(1-34)

where
=

(1-35)

and
=

(1-36)

The critical value of the stability number was determined to be 13.56. This number and its critical value provide a necessary and sufficient condition for predicting the onset of 20

instability in both oil-wet and water-wet porous media. If in a given cylindrical system the computed value of the stability number exceeds the critical value of 13.56, then the displacement will be unstable. In addition, the magnitude of the stability number provides quantitative information regarding the severity of instability. The higher the stability number, the more severe is the degree of instability. If the stability number for a displacement is less than the critical value, then the displacement will be stable.

1.4.3. Lake number

Beginning with a partial differential equation describing two-phase flow, Rapoport and Leas identified a group of variables whose value, it was argued, indicates when a flow is stabilized, i.e., becomes independent of capillary pressure gradient effects. The dimensional group they identified is (Lv), where v is the speed of the injected phase and is its viscosity. In the literature, the Rapoport and Leas (RL) scaling group has often been used to select the rate of water flood. Lake defined a dimensionless group that included the RL criterion, Lv, interfacial tension, contact angle, end-point relative permeability to the water phase and geometrical parameters of the porous medium. The Lake number, NL, is defined as
= cos

(1-37)

The appeal of the dimensionless Lake number is that it incorporates many of the important parameters pertaining to immiscible two-phase flow, including the pore geometry factor, K (it employs the inverse of this factor). Its apparent lack of success with the present set of data may be due to the use of a constant pressure drop boundary condition instead of constant flow rate conditions. Lake found that a NL of 3 corresponded to a scaling group of 1 in Rapoport and Leas data. The advantage of the dimensionless number is that it also includes permeability of the rock. For small NL, capillary pressure (Pc) will cause shock waves to spread out. The limit of rate criteria above which relative permeability is negligibly affected by capillary forces in a1-D wateroil displacement, when the core is water wet, is NL equal to 3.

21

1.4.4. Capillary end-effect number

The capillary end effect can be represented by a dimensionless flow parameter, Nc,end, the capillary end-effect number (Mohanty and Miller, 1991). It interpreted as the ratio of the capillary force to the viscous force at the system or macroscopic scale. Nc,end is the ratio of a characteristic capillary pressure, Pci, to the viscous pressure drop across the core, p; ie.,
, = //

(1-38)

p can be the initial pressure drop and Pci the capillary pressure at the initial saturation. Pci is frequently unavailable and thus is approximated by /, where k is a single-phase macroscopic capillary number, Ncm, introduced by Islam and Bentsen. This parameter has a critical value range in which it affects the JBN-derived relative permeability. For Nc,end >0.1, as Nc,end increases, both oil and water relative permeabilities decrease. When Nc,end < 0.1, it does not affect relative permeability.

permeability of the core. This dimensionless number is similar to the inverse of the

1.4.5. Capillary-to-gravity number

Saedi (2007) defined capillary-gravity number as the ratio of the threshold capillary pressure to the pressure differential created by centrifugation in the core sample for centrifugal displacement:
=

(1-39)

Larger values of capillary-gravity number imply a larger capillary end effect. While Ncg is directly related to threshold capillary pressure, it is implicitly a function of permeability. Hagoort (1980) used the capillary to gravity number and showed that for Ncg > 103 the capillary effect is important and needs to be taken into account. Saeedi (2007) performed a large number of centrifuge experiments and numerical investigation and stated that if the capillary to gravity number, Ncg, is maintained in order of 102 or smaller, the capillary endeffect becomes negligible. 22

1.4.6. Gravity number

It is the ratio of gravity forces to viscous forces.
=

(1-40)

where, v is the Darcy velocity, is the density difference between the displacing and displaced phases, is the viscosity diference, g is the permeability. The gravity number being a combination of the Bond and capillary numbers appears to be a better basis for comparison of laboratory and field data as it includes the gravity and viscous forces thereby indicating the relative effects of drainage versus displacement. gravity, k is the medium

1.5. Dimensionless Groups for Fluid Flow in Fractured Porous Medium

1.5.1. Fracture Capillary Number
Putra (1999) proposed parameters that affect the viscous and capillary forces can be grouped into a Fracture Capillary Number (FCN) equation. In naturally fractured reservoirs, fluid displacement in fracture network occurs due to its higher conductivity compared to the matrix while an exchange of fluids occurs between matrix and fracture system. The fluid transfer process is controlled by flow of water under naturally imposed pressure gradients (viscous force) and the spontaneous movement of water into the matrix under capillary forces (imbibition). The parameters that affect this process can be grouped into a dimensionless equation, called the fracture capillary number. This equation is used to determine the critical injection rate and is defined as the ratio between the viscous force that is parallel to the fracture direction and the capillary force that is perpendicular to the fracture direction. The viscous force is defined as a function of water velocity, water viscosity and fracture volume and is assumed to occur only in the fracture. The capillary force that occurs only in the matrix is defined as a function of interfacial tension, contact angle and matrix volume. Thus, the fracture capillary number can be written below: 23

= cos

(1-41)

where v is the velocity of the injected fluid in the fracture, defined as

, where qinj is the

volumetric injection rate and Af is the fracture cross-sectional area. Af is the fracture area, where the viscous force occurs, w is the water viscosity, is the interfacial tension and is the contact angle. cos can be substituted with
,

capillary pressure at Swi, km is the matrix permeability, m is the matrix porosity, and J(Swi) is the J-function at Swi. Am is the matrix area, where the capillary force.

where Pc,max is the maximum

Rearranging and substituting all those parameters into above equation, it can be rewritten as follows:
= ,

(1-42)

1.5.2. Modified Diffusive Capillary Number

Stubos and Poulou (1999) introduced a modified diffusive capillary number by taking the ratio of diffusion to capillary gradient driven rates. Their work mainly focused on drying of porous media with existence of two-phase.
=

(1-43)

where Deff is an effective diffusion coefficient through the porous matrix, Ce is the equilibrium concentration of vapor in the gas phase under the prevailing conditions, l is the liquid phase viscosity, l is the liquid phase density, km is the matrix permeability, is the porosity, is the interfacial tension.

1.5.3. Fracture Diffusion Index

24

Trivedi and Babadagli (2006) proposed a dimensionless group, Fracture Diffusion Index (FDI), as a ratio of viscous forces in the fracture and diffusion forces in the matrix using the analogy to capillary imbibition controlled immiscible displacement in fractured porous medium.
=

(1-44)

where v is the volumetric injection rate of the diffusing phase, kf is fracture permeability, km is the matrix permeability, s is the density of injected (solvent) phase, o is the density of displaced (oil) phase and D is the diffusion coefficient between oil and solvent. f() is the function of wettability. It is obtained from the slope of the straight line when the amount imbibed water is plotted against the square root of time for spontaneous imbibition test of the samples. Viscous flow is dominant in the fracture while the diffusional process accounts for the masstransport from fracture into the matrix. When applied at high solvent injection rate, most of the recovery is from the fracture through viscous flow and recovery from matrix is low due to poor diffusion into matrix. This yields a higher FDI. In the case of lower injection rates, the diffusion dominates the recovery and hence resulting into lower fracture diffusion index. Hence, higher FDI indicates a faster recovery with more solvent injection and presumably less ultimate recovery from the matrix. On the other hand, a low value of the FDI is an indication of slow but more efficient recovery. This might yield higher ultimate recoveries from the matrix, the process being slow due to low injection rate. Therefore, one needs to define a critical rate or critical FDI as a more universal indicator of the efficiency of the process.

1.5.4. Matrix-Fracture Diffusion Number

Trivedi and Babadagli (2008) proposed new group incorporating the matrix-fracture diffusion transfer to scale the miscible displacement in fractured porous media based on the laboratory experiments. It was named Matrix-Fracture Diffusion Number (NM-FD) and defined as:
= =

(1-45)

25

where o is the solute viscosity, uT is the flow velocity, b is the fracture width, kf is the fracture permeability, is the density difference, g is the gravity force, Dm is the Effective diffusion coefficient in matrix for matrix-fracture, L is the length of fracture, m is the matrix porosity, r is the matrix width, Pe is the Peclet number defined as:
=

(1-46)

Arm is the aspect ratio for matrix defined as:

=

(1-47)

and Ng is the gravity number defined as:

=

(1-48)

The expression for the (NM-FD) proposed is similar some extent to fracture capillary number, Nf,Ca, where the capillary forces are replaced by diffusive forces in the matrix with addition of gravity factor.

1.5.5. Dimensionless Time

Mattax and Kyte presented an equation for scaling of imbibition data under the following conditions: (1) the sample shape and boundary conditions are identical (2) the oil/water viscosity ratio is duplicated (3) gravity effects can be neglected (4) initial fluid distributions are duplicated (5) the capillary pressure functions arc directly proportional and (6) the relative permeability functions are the same. This group expresses the ratio of capillary force to viscous resistance, which is a form of inverse capillary number.

26

k t = t L

(1-49)

where t is the time, k is the permeability, is the fractional porosity, is the interfacial tension, w is the water viscosity, L is a characteristic length. The characteristic length was not further defined by Mattax and kyte. Ma et al. modified the Mattax and Kyte scaling law to define a dimensionless time for countercurrent imbibition in which gravity has no effect and the viscosities are finite:
1 =

(1-50)

Lc is the characteristic length determined by the size, shape, and boundary conditions of the sample defined by Ma et al. as:
= 1

(1-51)

1 =

(1-52)

where V is the volume of the sample, A is the area open to flow, l is the distance from the open surface to the no-flow boundary and the sum is over all open surfaces of the block. Ma et al. found that plots of the fraction of recoverable oil produced as a function of the dimensionless time, Eq. 1, for cores of different size with different fluid viscosities, boundary conditions, and interfacial tensions all lay approximately on the same universal curve. Zhou et al. used another equation to scale imbibition experiments in water-wet diatomite where the viscosity ratio was varied by four orders of magnitude:
= 1

(1-53)

27

where r*(=kr*/) is a characteristic mobility and M*(=rw * /ro * ) is a mobility ratio. They used endpoint relative permeabilities to calculate r* and M*. Eqs. (1-50) and (1-53) were developed for strongly water-wet rocks, where it is assumed that the oil/water contact angle is close to zero. To account for larger oil/water contact angles in mixed-wet media, Gupta and Civan and Cil et al. presented the following expression based on Eq. (1-50):
cos 1 =

(1-54)

Zhou et al. used a similar approach by defining an apparent dynamic advancing contact angle, AD, to match their experimental results. They defined this angle as:
= cos , 0.5 0.5

(1-55)

where the subscript vsww refers to very strongly water-wet conditions and tD(0.5) is defined as the dimensionless time, Eq. 1, for one half of the total recovery. This gives a dimensionless time given by Eq. 5 with =AD. Xie and Morrow tested Eq. 1 on 32 weakly water-wet Berea sandstone samples. They suggested that when capillary forces are sufficiently small, gravity segregation will make a significant contribution to oil recovery; therefore, this force must be included in scaling laws for weakly water-wet rocks:
= +

(1-56)

where Pc is a representative imbibition capillary pressure proportional to /K/, f() is a wettability factor, and LH is the vertical height of the sample.

28

1.6. New Dimensionless Groups as Combination of Dimensionless Numbers

1.6.1. Grattoni et al. Number (2001)
A new dimensionless group N is proposed by Grattoni et al. (2001) which is a linear combination of the three-phase Bond number and capillary number. This group represents the relative strength of the combined effects of gravity and viscous forces to capillary forces. This new group can be expressed as:
= +

(1-57)

where d is the viscosity of displaced phase and g is the viscosity of displacing phase. They developed a linear relationship between this new group and the total recovery based on the experimental investigations to include the pore scale effects.
= [ + ]

(1-58)

Then m and A are obtained from linear regression of R versus NB close to the end of the experiment and R versus NC at the beginning of the experiment.
= =

(1-59)

(1-60)

1.6.2. Kulkarni and Rao Number (2006)

Kulkarni and Rao (2006) presented the effect of major dimensionless groups on the nal recovery based on various miscible and immiscible gas assisted gravity drainage eld data and laboratory experimental data. The NG versus oil recovery plot did not yield as good a correlation. However, the individual plots of NC and NB versus recovery resulted in good 29

correlations. Therefore, it was hypothesized that there is some other important mechanistic parameter that is not well represented in the gravity number, and a mathematical combination of the NC, NB and NG groups with that mechanistic parameter should yield an improved correlation parameter. Literate review suggested that two ratios: density and viscosity (gas to oil), were important for gravity drainage flow. The density ratio was factored into a newly defined dimensionless group. They have termed this new dimensionless group as "Gravity Drainage Number" of NGD.
= + +

(1-61)

1.6.3. Rostami et al. Number (2009)

Rostami et al. (2009) conducted a number of forced gravity drainage experiments. It was found that neither the capillary number, nor the Bond number is sufficient to obtain a satisfactory correlation with recovery explain the full range of behaviors observed. In order to account for the effect of viscosity ratio (r = g/o) and proper ratio of the Bond to capillary number, a combined dimensionless group, NCo, is proposed:
=

(1-62)

where subscript Co stands for the term combined, and the parameters A and B are the scaling factors. They found that the values of A = B = 0.5 provide a good correlation. The small values of exponent suggest that the effect of the capillary number and the viscosity ratio is less than the importance of the Bond number. It is found that if recovery at a higher PV-injection was used, other values of A and B are needed to obtain a good match. Nevertheless, in all the cases the A and B exponents were smaller than 1.

1.6.4. Rostami et al. Number (2010)

In addition, Rostami et al. (2010) proposed a combined dimensionless group to generalize the estimation of remaining and residual oil saturations. Examination of residual oil saturations as a function of Bond and capillary number suggests that none of them individually can predict the remaining oil saturation. According to their experimental results, residual oil 30

saturation has an opposite relation to the Bond and capillary number. It is also demonstrated that the effect of capillary number is less than the importance of Bond Number. Therefore, the existence of the scaling factor is necessary.
=

(1-63)

where A is the scaling factor. This dimensionless group successfully combines the effect of main forces exist in forced gravity drainage flow.

1.7. Dimensionless Groups for Scaling EOR Processes

Shook et al. (1992) presented dimensionless scaling groups for the waterflood applicable to represent the two phase flow through homogeneous 2-dimensional Cartesian dipping reservoir. Effective Aspect ratio:

(1-64)

Mobility ratio:
=

(1-65)

Dip Angle group:

= tan

(1-66)

Buoyancy number:
=

cos

(1-67)

31

Capillary number:
=

(1-68)

Where H is the height of reservoir, L is the distance between injection and production wells, kx is the horizontal permeability, kz is the vertical permeability, koro is the endpoint relative permeability of oil, korw is the endpoint relative permeability of water, Mwo mobility ratio (water), uT is the fluid velocity, is the dip angle, is the difference between oil and water densities, o r2 is the endpoint mobility of oil, o is the viscosity of oil, w is the viscosity of water, g is the acceleration because of gravity, is the interfacial tension, is the porosity. Gharbi et al. (1998) found that there are eight independent dimensionless scaling groups that describe miscible displacement of oil by a solvent in two-dimensional, homogeneous, anisotropic cross section with constant porosity and dip angle. The eight dimensionless scaling groups describing the displacement are: Dispersion number:

(1-69)

Effective Aspect ratio:

(1-70)

Longitudinal Peclet number:

= /

(1-71)

Damkholer number:
=

(1-72)

32

Viscosity ratio:
=

(1-73)

Dip angle number:

= (1-74)

Gravity number:
=

(1-75)

Aspect ratio:
=

(1-76)

where KT and KL are the transverse and longitudinal dispersion coefficient, is the chemical rate constant, kx and kz are the absolute permeability in the x- and in the z-directions, H is the reservoir thickness, L is the characteristic system length, dip angle. If the molecular diffusion in the dispersion coefficients is negligible, then the longitudinal Peclet number and the dispersion number become:
= =

(1-77)

(1-78)

where L and T are the longitudinal and the transverse dispersivity, respectively.

33

Also, Gharbi (2002) made use of dimensionless groups for scaling of miscible ooding in a 2D heterogeneous anisotropic reservoir. They used several dimensionless scaling groups describing the reservoir heterogeneity presented by Li and Lake (1996). Global heterogeneity number (for log-normal)
= (1-79)

Local heterogeneity number (for log-normal)

=

(1-80)

Effective correlation length in x-direction

=

(1-81)

Effective correlation length in z-direction

=

(1-82)

Hurst exponent
(1-83)

where ln is the standard deviation of log-normal permeability field, Vp variance of permeability field, x and z are the correlation length in x-and z-directions. Wood et al. (2006, 2008) introduced ten dimensionless groups to describe CO2 ooding in a dipping waterooded reservoir, and used them for experimental design purposes to develop screening criteria applicable to Gulf Coast reservoirs. The dimensionless groups proposed by Shook (1992) served as the initial basis for the scaling of CO2 flooding. The relationship between the scaling groups and their effects on the overall performance of immiscible gas-driven gravity drainage were investigated by Jadhawar and Sarma (2008) 34

through the use of uncertainty assessment techniques. They used dimensionless scaling groups of previous studies for dipping reservoirs (Gharbi et al., 1998; Shook et al., 1992) and modified them to implement sensitivity analysis for the horizontal type reservoir under investigation. To consider the GAGD-EOR process dependent constant pressures condition of the injection wells and the production wells, and the highly compressible nature of the injectant (CO2), they modified gravity number to make it applicable to the horizontal type reservoir. Gravity number (based on the gas injection rate):
=

(1-84)

Gravity number (based on the gas injection and oil production pressures):
=

(1-85)

1.8. Application
The GAGD process extends the highly successful gravity stable gas floods in pinnacle reefs and dipping reservoirs to horizontal type reservoirs. To allow for scalability of the laboratory experiments, the reproduction of the various multiphase mechanisms and fluid dynamics, which have been found to be influential in the success of the gravity stable gas floods is crucial. Literature reviews of multiphase mechanics and fluid dynamics, suggests that dimensionless characterization of flood parameters to generate analogous field scale multiphase processes into the laboratory, is one of the most effective and preferred scaling tools (Kulkarni, 2004). Dimensional analysis becomes especially necessary for better understanding and performance characterization of novel processes like the GAGD. The dependant and independent variables used in this analysis are shown in Table 1.3 along with their fundamental dimensions. The various dimensionless groups obtained after the analysis are included as Table 1.4. It is important to note that the Buckingham-Pi analysis does not rank the dimensionless groups obtained in order of relative importance as controlling variables of the process. Literature

35

review, experimentation and inspectional analysis are required to determine the order of controlling groups of variable(s).
Table 1.3: Dependant and Independent Variables used for Buckingham-Pi Analysis

Inspectional analysis of the various dimensionless groups obtained via Buckingham-Pi analysis suggested the use of five dimensionless groups for the detailed analysis of the GAGD process namely, Capillary, Bond, Gravity, Dombrowski-Brownell, and N (new Grattoni et al.s group).

Table 1.4: Dimensionless Groups Obtained Using Buckingham-Pi Analysis

36

Nine commercial gas gravity drainage field applications were considered for the identification and characterization of various multiphase mechanisms, fluid dynamics and calculation of the range of various dimensionless groups probable in the GAGD process. The protocol followed to facilitate their calculation is included as Figure 1.1. Calculation of these dimensionless numbers for field projects involved the use of various well logs (for thickness, net-to-gross values, OWC, GOC and grain size), field maps (for Darcy velocity), grain size classification systems (for Bond number), production / injection data (for New Grattoni et al.s group), bottom hole pressure survey plots (for PVT simulations), compositions of injected / produced fluids (for PVT simulations), and PVT simulations (for fluid properties predictions).

Figure 1.1: Protocol for Calculation of Dimensionless Groups for Field.

All the experiments are designed to mimic or resemble the multiphase mechanisms operative in the field processes by keeping synonymous dimensionless values obtained from the field projects. The grain size of the glass beads and fluid properties in the experimental model are selected to generate similar values of the selected dimensionless numbers. Variables in the dimensionless numbers were varied to capture the entire spectrum of dimensionless number values obtained from the field projects. From Bond number equation we can see that, NB is directly proportional to the absolute permeability of the porous medium and the density 37

difference between the reservoir fluids. Absolute permeability of a consolidated porous medium is a strong function of the grain diameter and is given by the Carman-Kozeny (Carman, 1937) equation:
=
721

(1-86)

In above equation, DP is the grain diameter, is the tortuosity and is the porosity of the porous medium. In order to obtain favorable and realistic Bond numbers, fluid-fluid interaction parameters (interfacial tension) are also important. On the other hand, different capillary numbers were obtained by varying the fluid-fluid systems, for example DecaneCO2 and Paraffin-CO2 in the physical model. However, the ranges of capillary number obtained through selection of different fluid-fluid system were not as large in comparison to the ranges obtained through selection of different gas injection rates. Therefore different gas injection rates, using a mass flow controller, were used to generate capillary numbers of various orders of magnitude. The operating rages of these field projects were duplicated in the laboratory for 1-D corefloods and 2-D physical model experiments, by proper fluid and operating condition selections. The values of the dimensionless groups obtained for the laboratory experiments have also been included in Table 1.5.
Table 1.5: Dimensionless Number Ranges Obtained for Field Applications and Laboratory Studies

38

In order to scale-up the laboratory run time to a given prototype field, the following dimensionless time expression was used. The expression for the dimensionless time (td) for gravity drainage processes was obtained from the literature (Miguel et al., 2004) and is expressed as:

(1-87)

where k is the absolute permeability of the porous medium, Koro is the end-point relative permeability to oil, is the density difference between the displaced phase and the displacing phase, g is the acceleration due to gravity, h is the thickness of the porous medium, is the porosity, o is the oil viscosity, Sor and Swi are the residual oil and connate water saturation respectively, td is the dimensionless time, and t is real time. Above equation enables the scale-up of the run time (in minutes) in the experimental models to time required in a prototype reservoir to reach similar recoveries. The results obtained from the physical model and immiscible core flood experiments are compared with data obtained from the gravity drainage field projects. For effective comparisons, as well as to account for the relative variations of the Bond and Capillary numbers in each of these floods, a single comparison parameter is required. The gravity number is a combination of Bond and capillary numbers, and incorporates the relative variations of the major reservoir forces, namely the gravity, capillary and viscous forces. Therefore, the Gravity number appeared to be more appropriate for the comparison of laboratory and field data. This is very encouraging, since the data for this comparison are obtained from vastly varied sources, such as from the atmospheric pressure, homogeneous 2-D sand packs, to the highly heterogeneous and high-pressure field flood projects. These findings indicate that the performance of the GAGD process appears to be well characterized by the use of the gravity number.

39

1.9. References
Babadagli, T., 2002. Scaling capillary imbibition during static thermal and dynamic fracture ow conditions. J. Pet. Sci. Eng. 33, 223. Babadagli, T., Ershaghi, I., 1993. Improved modeling of oil/water ow in naturally fractured reservoirs using effective fracture relative permeabilities. SPE 26076, SPE Western Regional Meeting, Anchorage, Alaska, May 2628, 1993, pp. 463476. Bear, J., 1972. Dynamics of Fluids in Porous Media; Elsevier Publishing Co., Inc.: New York, 665727. Buckingham, E., 1914. On Physically Similar Systems: Illustrations of the Use of Dimensional Equations, Phys. Rev., vol. 4, No. 4, pp. 345376. Carman, P. C., 1937. Fluid Flow through Granular Beds, Trans. Inst. Chem. Engrs., London, 15-150. Cil, M., Reis, J. C., Miller, M. A., Misra, D., 1998. An Examination of Countercurrent Capillary Imbibition Recovery From Single Matrix Blocks and Recovery Predictions by Analytical Matrix/Fracture Transfer Functions, paper SPE 49005 presented at the SPE Annual Technical Conference and Exhibition, New Orleans, 2730 September. Craig, F.F., Sanderlin, J.L. and Moore, D.W. 1957. A Laboratory Study of Gravity Segregation in Frontal Drives. SPE 676-G, Petroleum Transactions, AIME, 210: 275282. Djuro, N., 2002. Numerical Reservoir Characterization Using Dimensionless Scale Numbers with Application In Upscaling. Phd Dissertation, Louisiana State University. Dombrowski, H.S., Brownell, L.E., 1954. Residual equilibrium saturation of porous media. Ind. Eng. Chem. 46, 12071219. Geertsma, J., Croes, G.A. and Schwarz, N. 1956. Theory of Dimensionally Scaled Models of Petroleum Reservoirs. Pet. Tran. AIME, 207. Gharbi Ridha, B.C. 2002. Dimensionally Scaled Miscible Displacements in Heterogeneous Permeable Media. Transport in Porous Media, 48: 271-290. Gharbi, R., Peters, E. and Elkamel, A. 1998. Scaling Miscible Fluid Displacements in Porous Media. Energy & Fuels, 12: 801-811. Grattoni, C.A., Jing, X.D. and Dawe, R.A. 2001. Dimensionless groups for three-phase gravity drainage flow in porous media. Journal of Petroleum Science and Engineering, 29(1): 53-65. Greenkorn, R.A. 1964. Flow Models and Scaling Laws for Flow Through Porous Media. Industrial and Engineering Chemistry, 56(3): 32-37. 40

Gupta, A. and Civan, F., 1994. An Improved Model for Laboratory Measurement of Matrix to Fracture Transfer Function Parameters in Immiscible Displacement, paper SPE 28929 presented at the SPE Annual Technical Conference and Exhibition, New Orleans, 2528 September. Hagoort, J. 1980. Oil Recovery by Gravity Drainage SPE Journal (SPE 7424): 139-150. Handy, L.L., 1969. Determination of effective capillary pressure for porous media from imbibition data. Trans. AIME 75, 219. Islam, M.R. and Bentsen, R.G.: "Effect of Different Parameters on Two-Phase Relative Permeability," AOSTRA J. Research (1987) 3, 69-90. Jadhawar, P.S. and Sarma, H.K. 2008. Numerical Simulation and Sensitivity Analysis of Gas-Oil Gravity Drainage Process of Enhanced Oil Recovery. CIPC-006 Canadian International Petroleum Conference, Calgary, Canada, June 17-19. Kulkarni, M. M., 2004. Multiphase Mechanism and Fluid dynamics in Gas Injection Enhanced Oil Recovery Processes, LSU Ph. D. Thesis. Kulkarni, M.M. and Rao, D.N. 2006. Characterization of Operative Mechanisms in Gravity Drainage Field Projects through Dimensional Analysis. SPE 103230, presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, September 2417. Lake, L.W., 1989. Enhanced Oil Recovery. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Leverett, M.C., Lewis, W.B. and True, M.E. 1942. Dimensionsal-model studies of oil-field behaviour. Trans. AIME, 146: 175-193. Lewis, J.O. 1944. Gravity Drainage in Oil fields. AIME Petroleum Transactions, 155(133). Li, D. and Lake, L.W. 1995. Scaling Fluid Flow Through Heterogeneous Permeable Media. SPE 26648, SPE Advanced Technology Series, 3(1): 188-197. Lozada, D. and Farouq Ali , S.M. 1987. New Sets of Scaling Criteria for partial Equilibrium Immiscible carbon Dioxide Drive. Journal of Canadian Petroleum Technology: 393-411. Ma, S., Morrow, N.R., and Zhang, X., 1997. Generalized Scaling of Spontaneous Imbibition Data for Strongly Water-Wet Systems, J. of Petroleum Science and Engineering, 18, 165. Mattax, C.C., Kyte, J.R., 1962. Imbibition Oil Recovery From Fractured, Water-Drive Reservoir, SPEJ, 177; Trans., AIME, 225. Miguel, N. H., Miller, M. A. Sepehrnoori, K., 2004. Scaling Parameters for Characterizing Gravity Drainage in Naturally Fractured Reservoirs, SPE IPCM.

41

Mohanty, K.K., Miller, A.E., 1991. Factors influencing unsteady relative permeability of a mixed-wet reservoir rock. Soc. Pet. Eng. Form. Eval. 6, 349358. Morrow, N.R., Chatzis, I., 1984. Correlation of capillary number relationships for sandstone. SPE 10114, 59th SPE Annu. Tech. Conf. Exhibit., Dallas, TX. Perkins, F.M. and Collins, R.E. 1960. Scaling laws for laboratory flow models of oil reservoirs. SPE 1487-G, Journal of Petroleum technology: 69-71. Peters, E.J. and Flock D.L., 1981. The Onset of Instability During Two-Phase Immiscible Displacement in Porous Media. Soc. Petrol. Eng. J., April, 249258. Rao, D.N., Ayirala, S.C., Kulkarni, M.M. and Sharma, A.P. 2004. Development of Gas Assisted Gravity Drainage (GAGD) Process for Improved Light Oil Recovery. SPE 89357, presented at the 2004 SPE/DOE Fourteenth Symposium on Improved Oil Recovery Tulsa, Oklahoma, U.S.A., April 1721. Rapoport, L.A. 1955. Scaling laws for use in the design and operation of water-oil flow model. Trans. AIME, 204: 143-150. Rapoport, L.A. and Leas W.J., 1953. Properties of Linear Waterfloods. Trans. AIME, 198, 139148. Reynolds, O., 1883. An Experimental Investigation of the Circumstances which Determine Whether the Motion of Water Shall Be Direct or Sinuous and of the Law of Resistance in Parallel Channels, Phil. Trans. R. Soc., vol. 174, pp. 935982. Rostami, B., Kharrat, R, Pooladi-Darvish, M., Ghotbi C., 2010. Identification of Fluid Dynamics in Forced Gravity Drainage Using Dimensionless Groups. Transp. Porous Media, 83, 725740. Rostami, B., Kharrat, R., Ghotbi, C., Tabatabaie, S.H., 2010. Gas-Oil Relative Permeability and Residual Oil Saturation as Related to Displacement Instability and Dimensionless Numbers. Oil & Gas Science and Technology Rev. IFP, 65 , 2, 299-313. Ruark, A.E., 1935. Inspectional Analysis: A Method Which Supplements Dimensional Analysis, J. Elisha Mitchell Sci. Soc., 51, 127132. Shook, M., Li, D. and Lake, L.W. 1992. Scaling Immiscible Flow through Permeable Media by Inspectional Analysis. In-Situ 4: 311-349. Stubos, A.K., Poulou, S., 1999. Oil recovery potential from fractured reservoirs by mass transfer processes. SPE 56415, paper presented at SPE Annual Technical Conference and Exhibition, Houston, TX. October. Trivedi, J. and Babadagli, T. 2008. Scaling miscible displacement in fractured porous media using dimensionless groups. Journal of Petroleum Science and Engineering. 42

Trivedi, J.J., Babadagli, T., 2006. Efciency of miscible displacement in fractured porous media. SPE 100411.Western Regional/AAPG Pacic Section/GSA Cordilleran Section Joint Meeting, Anchorage, Alaska, 810 May, 2006. Trivedi, J.J., Babadagli, T., 2007. Efciency of diffusion controlled miscible displacement in fractured porous media. Transp. Porous Media in print. White F., M., 1994. Fluid Mechanics, McGraw-Hill, 4th Edition. Wood, D.J., Lake, L.W., Johns, R.T. and Nunez, V. 2008. A Screening Model for CO2 flooding and storage in Gulf Coast Reservoirs based on Dimensionless Groups. SPE Reservoir Evaluation and Engineering, 11(3): 513-520. Wood, D.J., Lake, L.W., Johns, R.T., 2006. A screening model for co2 ooding and storage in gulf coast reservoirs based on dimensionless groups. SPE 10021, SPE/DOE Symposium Improved Oil Recovery, Tulsa, OK. April 2006. Xie, X., Morrow, N.R., 2000. Oil Recovery by Spontaneous Imbibition From Weakly WaterWet Rocks, paper SCA 2000-27 presented at the Intl. Symposium of the Soc. of Core Analysts, Abu Dhabi, UAE. Zhou, X., Morrow, N.R., and Ma, S., 2000. Interrelationship of Wettability, Initial Water Saturation, Aging Time, and Oil Recovery by Spontaneous Imbibition and Waterflooding, SPEJ, 199.

43