Applicability of Different Formulas For Effective Permeability

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 34

Applicability of different formulas for effective permeability of

a fracture network into a reservoir model

Bachelor thesis in physics

By Thom Postma
Student number: S2057573

First supervisor: prof. dr. Rien Herber


Second supervisor: prof. dr. Harro Meijer
Daily supervisor: Alex Daniilidis

Energy and Sustainability Research Institute Groningen: Geo-energy


University of Groningen

July 2014

1
Abstract

Different formulas for the effective permeability of a fracture have been assessed. This was done by
doing geothermal simulations in PetraSim. The simulations consisted of the injection of water at a rate of
33 kg/s and a temperature of 10 °C into a 10x10x10m reservoir with a uniformly spaced fracture network
inside. Subsequently this water was being produced at a production well at another location. The
different formulas: , and were used to calculate the effective
permeability of a single fracture. Simulation results showed a difference in flow rate at the production
well for k1 and k2 caused by the fact that k2 can only be used to calculate single fracture permeability
when H0 equals h, in that case k2 becomes k1. Formula k1 was found to be useful for discrete fracture
network models, in which fractures are explicitly defined. It was found that an alternative way of using k2
is to calculate the average effective permeability over a cell consisting of fracture and matrix. In that
way, k2 becomes useful for large scale reservoir models since fractures do not have to be defined. A
disadvantage of this, is that it becomes less realistic since flow will be averaged over a cell comprising
fracture and matrix. Formula k3 was found to be useful for predicting the effective permeability of a
fracture after a stress induced aperture change, however simulation results depended strongly on the
value of k0 used.

2
Contents
1. Introduction ........................................................................................................................................... 5
2. Background............................................................................................................................................ 5
3. Methodology ......................................................................................................................................... 8
3.1 PetraSim and THOUGH2 ................................................................................................................ 8
3.2 Fracture modeling method............................................................................................................ 9
3.3 Fracture distributions .................................................................................................................... 9
3.4 Initial conditions .......................................................................................................................... 10
3.5 Power output............................................................................................................................... 10
3.6 Well parameters .......................................................................................................................... 10
3.7 Fracture permeability .................................................................................................................. 11
3.8 Materials...................................................................................................................................... 13
3.9 Mesh and layer ............................................................................................................................ 13
3.10 Solution controls ......................................................................................................................... 13
4. Simulation results ................................................................................................................................ 13
4.1 Temperature contours ................................................................................................................ 14
4.2 1x and 0y fractures ...................................................................................................................... 16
4.3 9x and 0y fractures ...................................................................................................................... 18
4.4 9x and 1y fractures ...................................................................................................................... 19
4.5 9x and 2y fractures ...................................................................................................................... 19
4.6 9x and 9y fractures ...................................................................................................................... 20
4.7 Overview different fracture distributions ................................................................................... 20
4.8 Different fracture porosity .......................................................................................................... 21
5. Discussion ............................................................................................................................................ 22
5.1 Flow behavior .............................................................................................................................. 22
5.2 Applicability of k1 ......................................................................................................................... 22
5.3 Applicability of k2 ......................................................................................................................... 23
5.4 Applicability of k3 ......................................................................................................................... 23
5.5 Recommendations....................................................................................................................... 23
6. Conclusion ........................................................................................................................................... 24
7. Acknowledgements ............................................................................................................................. 25

3
8. References ........................................................................................................................................... 26
9. Appendix A: Fracture permeability and productivity index calculated by the different formulas ..... 28
10. Appendix B: 1x and 0y simulation results ....................................................................................... 30
11. Appendix C: 9x and 1y simulation results ....................................................................................... 31
12. Appendix D: 9x and 2y simulation results ....................................................................................... 32
13. Appendix E: 9x and 9y simulation results ........................................................................................ 33
14. Appendix F: 1x and 9y simulation results ........................................................................................ 34

4
1. Introduction
Fossil fuels are the main energy source worldwide. However fossil fuels have high CO2 emissions and
reserves are being depleted. Therefore renewable energy sources will be needed to replace fossil fuels.
Geothermal energy is a renewable energy source that makes use of the earth's heat. A way of extracting
geothermal energy is injection of water in an underground aquifer and producing the heated water at
another location. The Eemshaven is a possible location for geothermal energy in The Netherlands. Ideally
the water would be injected into the Rotliegend reservoir, however since this already contains the
Slochteren gas field this is not possible. Alternatively a fractured carbonate reservoir close to the top of
the Zechstein could be used for geothermal energy. This rock has a very low porosity, but contains
fractures due to its brittle deformation. These fractures can form a network which can potentially
facilitate fluid flow. The effective permeability describes the ability of a fluid to flow through rock.
Several researchers have described the effective permeability of a fracture network in different ways.
The aim of this thesis is to assess the formulas for effective permeability in a fracture network and to find
the formula that is most suitable for implementing into a reservoir model. Formulated as a research
question: What is the applicability of the different formulas for effective permeability of a fracture
network into a reservoir model. To find this out, a series of simulations will be done in PetraSim, a
graphical interface program for reservoir modeling. The output of these simulations will give an
indication of these formulas being realistic, and thus applicable for reservoir modeling.

2. Background
The power output of a geothermal doublet is defined by:

(1)

Where ρw is the water density, cw the specific heat of water, T0 the cold water temperature at the inlet,
T1 the initial aquifer temperature and Q the volume flow rate of water. The volume flow rate is given by
Darcy's law:
(2)

Where k is the effective permeability of the rock, µ the fluid's dynamic viscosity, A the cross-sectional
area, L the distance and Δp the pressure drop over the distance. The SI unit of permeability is m2, but
often the darcy (1 darcy=9.87*10-13 m2) is used as a practical unit. From this it can be seen that a high
permeable rock will have a high flow rate and thus a high power output. Therefore high permeable rocks
are favorable for geothermal projects.

Previous researchers have described the effective permeability of a fracture network on several different
ways. Sarkar et al. (2004), assumed that the fracture walls could be represented by two smooth parallel
plates which are separated by an aperture h. Furthermore they assumed that there is a constant static
pressure at the inlet and outlet of the fracture and that the pressure gradient through the fracture is
uniform. The analytical solution for the velocity through this fracture is [Wilkes, 1999]:

(3)

5
Where u(z) is the velocity in the x-direction and Pi and Po the pressures at the inlet and outlet of the
fracture. The flow rate through a fracture (figure 1) with width W and aperture h can then be found by
integrating equation (3) from z=0 to z=h:

(4)

Figure 1. Schematic of a fracture simplified by two parallel plates [Sarkar et al., 2004] .

Equation (4) is the so called “cubic law”. Combining Darcy’s law (2) and the cubic law (4), gives the
permeability of the fracture:

(5)

Multiplying this equation with the area A gives the transmissivity:

(6)

However since fracture walls aren’t smooth, an equivalent aperture heq should be used to replace h with
[Sarkar et al., 2004].

Roman et al. (2012), developed an elastic-plastic model that describes the permeability of a single
fracture. The model is subjected to elastic deformation as well as plastic deformation (unrecoverable
strains) as a result of a varying net confinement pressure Pc. Figure 2 shows a representative element
used in the model. It is assumed that a fracture can be described by a rigid impermeable block with an
idealized fractured surface, whose contact asperities can be deformed.

6
Figure 2. Schematic of the representative element of fractured media [Roman et al., 2012].

Furthermore it is assumed that the contact points of the asperities are semi-spherical with a radius of
curvature R. The flow through the fracture is idealized by a flow between two smooth parallel plates. The
flow rate is then given by a Poiseuille's flow:

(7)

Where p is the fluid pressure, µ the fluid viscosity and h the effective fracture aperture:

-α (8)

Where h0 is the initial aperture size and α the asperity's deformation. From equation (7) and Darcy’s law
(2) it can be derived that the flow through the representative element is:

(9)

Where H0 is the height of the porous media sample. Combining equations (7) and (9) gives the effective
permeability of the fractured media as [Roman et al., 2012]:

(10)

The previous two cases have both assumed that the flow in a fracture can be described by a flow
between two smooth parallel plates. However since real fractures aren’t smooth and parallel it is the
question if this assumption still holds. Tortuosity is a parameter of a fracture being tortuous, it is given by
the following formula:

(11)

Where L is the length of the fracture and Le the effective length of a streamline path through the fracture
[Carman, 1956]. Researchers showed that the cubic law (4) is still valid when a correction factor for
tortuosity is used [Zimmerman and Bodvarsson, 1995], or by using an effective aperture [Witherspoon et

7
al., 1980]. They also showed that permeability changes due to stress could be predicted, as long as the
fracture surface was not affected by shear movements or weathering [Witherspoon et al., 1980].
Researchers also found that applying higher stress and using larger samples increased deviation from the
cubic law (4). They showed that as a result of a changing stress, both the aperture and the degree of
asperity contact are changed, and therefore the permeability is changed [Raven and Gale, 1985].

Based on experimental results, Zhang et al. (2007), showed that there is a relationship between
permeability and stress induced aperture change. Their experiment was executed on a concrete block
with three to four fractures in it. The water flow rate in the z direction through this block was measured,
while biaxial loadings in the x and y directions were applied. Based on their results, cubic flow rate-stress
relationships and aperture-stress relationships were found. Combining these relationships gave the
relation between flow rate and aperture change:

(12)

Were C1 and C2 are constants and Δb is the aperture change:

(13)

Where b0 is the original aperture and b the aperture after the loading was applied. Combining Darcy's
law (2) and equation (12) gives the relationship between permeability and aperture change:

(14)

Where C0 is a constant, equal to the initial permeability k0 of the fracture before applying the load.
Equation (14) can now be expressed as:

(15)

This gives the formula for permeability as a result of stress induced aperture change [Zhang et al., 2007].

3. Methodology

3.1 PetraSim and THOUGH2


To assess the different formulas for effective permeability of a fracture network, a series of simulations
will be performed. The program used for these simulations is PetraSim [Thunderhead Engineering, 2012].
PetraSim is a graphical interface program used to visualize reservoir models. It is a pre- and
postprocessor for the TOUGH2 programming code. TOUGH2 is a numerical simulator for fluid and heat
flow in porous and fractured media [Pruess et al., 1999]. The working of TOUGH2 is based on the solving
of mass and energy balance equations for fluid and heat flow in multiphase and multicomponent
systems. It makes use of a multiphase extension of Darcy’s law, and diffusion to describe fluid advection,
while heat flow is described by conduction and convection. TOUGH2 provides several equation-of-state
(EOS) modules that are capable of simulating flow for various conditions. The module used for these
simulations will be EOS1, capable of modeling water in liquid, vapor or two-phase state.

8
3.2 Fracture modeling method
There are three approaches commonly used to describe fluid flow in fractured media. The first approach
is the dual-porosity/single-permeability model, in which the reservoir is idealized by equal size matrix
blocks separated by interconnected fractures. The matrix blocks are assumed to have a much larger
porosity than the fractures, while the fractures contain a much larger permeability than the matrix
blocks. It is therefore assumed that the global flow occurs only in the highly permeable fractures, while
matrix and fractures may exchange fluid locally [Warren and Root, 1963]. The second approach is the
dual-porosity/dual-permeability model, in which in contrast to the previous model there is also flow
inside the matrix [Pruess, 1983]. The third approach is the discrete fracture network (DFN) model, the
most recent method. It relies on three-dimensional spatial fracture planes which form an interconnected
fracture network. The matrix blocks are bounded by the fracture planes. The advantage of a DFN model
above the other two methods is that fractures can be presented as discrete features rather than
regularly spaced fractures among equal size matrix blocks [Gilman, 2003] .

The dual-porosity/single-permeability and dual-porosity/dual-permeability approaches don’t work in this


case, because the matrix which is an anhydrite has a very low porosity and permeability. For this
research a discrete fracture network model will be used, however the fractures will be uniformly
distributed so that the effects of different distributions can be examined. Before creating the model a
few characteristics of the network will have to be determined. At first there is the fracture density
describing the amount of fractures per area. Secondly there is the fracture orientation. Thirdly there is
fracture aperture. Finally there is fracture size. The fracture density is determined by the fracture
spacing. For the simulations, the fracture spacing was varied from 1 m to 10 m. Fracture apertures from
100 µm to 3 mm were used. The target reservoir volume has dimensions of 3 km by 0.8 km with a height
of 50 m. However since using this volume would require creating lots of fractures, a smaller control
volume of 10x10x10 m was used. The fractures are cutting through the whole volume, which means they
have a height of 10 m.

3.3 Fracture distributions

Figure 3. Fracture distributions: (1) uniform distribution in the x-direction direction (left), (2) uniform
distribution with perpendicular crossings (right).

9
Two different fracture distributions will be examined: (1) an uniform distribution of fractures orientated
in one direction, which is in the direction of a fault, and (2) an uniform distribution of fractures oriented
in one direction with other fractures crossing in the perpendicular direction. Throughout the simulations
the fracture density will be varied both in the x-direction as in the y-direction direction (figure 3).

3.4 Initial conditions


Before starting a simulation, PetraSim requires certain initial conditions to be entered. The reservoir has
an initial temperature of 60 °C. The steady state pressure in the reservoir can be calculated using:

(16)

Where ρ is the density of water, g the gravitational acceleration and h the depth of the reservoir. In this
case the reservoir is at 2 km depth, and the steady state pressure is around 2*107 Pa or 200 bar.

3.5 Power output


To see if the reservoir is suitable for geothermal energy projects a certain amount of power output is
required. In this simulation water of 10 °C is injected at a rate of 33 kg/s into the 60 °C reservoir.
Assuming that the outgoing flow has a temperature of 60 °C, this corresponds to a geothermal power of
almost 7 MW.

3.6 Well parameters


The way in which PetraSim defines a well is by creating a line in 3D space. This line does not represent a
true coupled well model, but is used to identify the cells that intersect the line. Subsequently source or
sink properties are assigned to each of these cells. To define an injection well, a fixed water injection rate
(kg/s) should be assigned, as well as a fixed enthalpy (J/kg). The injection rate will be 33 kg/s. The
enthalpy will be 61308.2 J/kg, corresponding to water at a temperature of 10 °C and pressure 2*107 Pa.
However from Darcy's law (2), it can be seen that in order to maintain the fixed flow rate of 33 kg/s the
pressure at the injection well will increase a lot if low fracture permeabilities are used and the pressure
at the production well stays around its initial value. Therefore the temperature of the injected water will
actually be lower when the pressure goes up. To define the production well, a “well on deliverability”
option will be used. This allows the production of the well to occur against a prescribed bottomhole
pressure. Assuming water in the reservoir is only present in the liquid phase, the mass production rate of
a producing cell will be:

(17)

Where µ is the viscosity of water, ρ the density of water, P the pore pressure of the cell, the
prescribed bottomhole pressure and PI the productivity index given by:

(18)

Where Δ is the layer thickness, k the permeability of the layer, re the element radius, rw the well radius
and s the skin factor [Pruess et al., 1999]. The value for permeability will depend on the fracture

10
characteristics and formula used to calculate it. The other values that will be used are: Δzl=50 m,
re/rw=1.0 and s=1.

3.7 Fracture permeability

Formula

Table 1. Formulas for effective permeability of a fracture.

The different formulas for effective permeability of a single fracture are shown in table 1. These are now
used to calculate the fracture permeability of different fracture networks. For k2, H0 equals the mean
fracture spacing. For k3, the values of k1 and k2 will be used for k0. In table 2 the fracture permeability
and productivity index as calculated by formula k1 are shown and the same is done for formula k2 in table
3. For formula k2 the permeability for x-direction and y-direction fractures was calculated separately,
because the x-direction and y-direction fracture spacing differs. It should also be noted that at cells were
x-direction and y-direction fractures cross, the highest permeability is taken for that cell. Finally the
permeability and productivity index as calculated by formula k3 are shown in table 4 for k0=k1 and in
table 5 for k0=k2. Complete tables with fracture permeability and productivity index for different
formulas and fracture distributions are shown in Appendix A.

Fracture aperture Permeability(m2) Productivity


index(m3)
3 mm 7.5*10-7 9.4*10-5
2 mm 3.3*10-7 4.1*10-5
1 mm 8.33*10-8 1.05*10-5
500 µm 2.08*10-8 2.62*10-6
300 µm 7.5*10-9 9.42*10-7
200 µm 3.33*10-9 4.18*10-7
100 µm 8.33*10-10 1.05*10-7

Table 2. Fracture permeability and productivity index for various fracture apertures calculated by using
formula k1.

11
Fracture X-direction Y-direction X-direction Y-direction X-direction Y-direction PI(m3)
aperture fractures fractures fracture fracture fracture fracture
spacing(m) spacing(m) permeabiltiy(m2) permeability(m2)
3 mm 1 0 10 X 2.3*10-10 x 2.9*10-8
3 mm 9 9 1 1 2.3*10-9 2.3*10-9 2.9*10-7
2 mm 1 0 10 X 6.67*10-11 x 8.38*10-9
2 mm 9 9 1 1 6.67*10-10 6.67*10-10 8.38*10-8
1 mm 1 0 10 X 8.33*10-12 x 1.05*10-9
1 mm 9 9 1 1 8.33*10-11 8.33*10-11 1.05*10-8
500 µm 1 0 10 x 1.04*10-12 x 1.31*10-10
500 µm 9 9 1 1 1.04*10-11 1.04*10-11 1.05*10-9
100 µm 1 0 10 x 8.33*10-15 x 1.05*10-12
100 µm 9 9 1 1 8.33*10-14 8.33*10-14 1.05*10-11

Table 3. Fracture permeability and productivity index for various fracture apertures calculated by using
formula k2.

Aperture change Permeability(m2) Productivity


index(m3)
From 1 mm to 3 mm 2.25*10-6 2.82*10-4
From 3 mm to 1 mm 2.78*10-8 3.49*10-6

Table 4. Fracture permeability and productivity index after an stress induced aperture change calculated
by using formula k3 and k0=k1.

Aperture Fractures (number and X-direction Y-direction X-direction Y-direction PI(m3)


change orientation) fracture fracture fracture fracture
spacing(m) spacing(m) permeability(m2) permeability(m2)
From 1 mm 1x and 0y 10 x 2.25*10-10 x 2.83*10-8
to 3 mm
From 3 mm 1x and 0y 10 x 8.5*10-12 x 1.1*10-9
to 1 mm
From 1 mm 9x and 9y 1 1 2.25*10-9 2.25*10-9 2.83*10-7
to 3 mm
From 3 mm 9x and 9y 1 1 8.5*10-11 8.5*10-11 1.1*10-8
to 1 mm

Table 5. Fracture permeability and productivity index after an stress induced aperture change calculated
by using formula k3 and k0=k2.

12
3.8 Materials
There are basically two types of materials involved in this case. The first is the matrix, which has the
properties of an anhydrite. Anhydrite has porosity ranging from 1 to 3 percent and permeability ranging
from 1*10-21m2 to 1*10-16m2 [Christiansen and Howarth, 1995]. Because these values are so low, for
most simulations the porosity and permeability were taken to be zero. Some extra simulations were also
carried out to examine the effect when a porosity of 2 percent and permeability of 1*10-17m2(0.01mD)
are used. Furthermore anhydrite has a density of 2830 kg/m3, a wet heat conductivity of 4.00 W/m K and
a specific heat of 830 J/kg K [Doddema, 2012]. The second material is the fracture which has a porosity of
0.5 and permeability as calculated by one of the formulas. It has the same density, wet heat conductivity
and specific heat as the matrix.

3.9 Mesh and layer


For the simulations a two dimensional model will be used consisting of one z layer and 19 x and 19 y
divisions.

3.10 Solution controls


The solution controls menu in PetraSim allows specifying simulation parameters. For all simulations a
start time of 0 s and end time of 40 years is chosen. An initial time step of 100 s is taken and the
maximum number of time steps is set to 200. The automatic time step adjustment option is selected,
which means that the initial time step is only used for the first time step and the program will use
suitable time steps ahead, these time steps can therefore also be larger than the small initial time step.

4. Simulation results
The simulations started with injection of water at a rate of 33kg/s into the 10x10x10m reservoir.
Simulations showed that flow between injection and production well, only occurred when both wells
were connected via fractures, if the anhydrite permeability was taken to be zero. No flow occurred
between fractures that were not connected as there is no flow possible through the impermeable
matrix. This means that fracture crossings are required for multiple fractures to have effect. Therefore to
always have flow, the injection and production wells were placed at both ends of the middle x-direction
fracture.

Some extra simulations were carried out to examine the effect if the anhydrite has a permeability of
1*10-17m2 and porosity of 2 percent. In this case the injection well was removed from a fracture and
placed in the matrix. It was found that there is still no flow going from the injection well to the
production well if they are not connected by fractures when formula k1 and fracture aperture 3mm were
used, with only one x-direction fracture. Even when placing the injection well as close as 0.5 m to the
fracture no flow occurred. The same simulation was done with nine x-direction and nine y-direction
fractures and showed a flow of 1.5 kg/s at the production well and a pressure of 1.2*109 Pa at the
injection well. This simulation ended with an error occurring after 2.15 seconds. The matrix porosity and
permeability were taken to be zero for all other simulations, because these are very small for anhydrite.

In this chapter the simulation results will be shown. While the fracture distribution was being held
constant, the fracture aperture was being varied. For each simulation the flow rate at the production

13
well and the pressure at the injection well at the end of the simulation are given. Pressure at the
production well is not given since it always stayed near the initial reservoir pressure. Furthermore it is
shown if a simulation error occurred. This is usually caused by a too high pressure near the injection well,
which is caused by the fact that the pressure at the injection well has to go up very high to ensure a large
enough pressure drop between injection and production well, to maintain a flow rate of 33 kg/s. In
reality a rock will crack after a certain pressure is being applied, however PetraSim gives an error and
ends the simulation in that case. Finally the simulation time is also given.

4.1 Temperature contours


Some temperature contours are shown below to get an understanding of the flow behavior in the
reservoir.

Figure 4. Temperature contours from above for one x-direction and zero y-direction fractures after 50
hrs.

For one x-direction and zero y-direction fractures all the flow is going through the single fracture and the
cold water is cooling the surrounding rock.

14
Figure 5. Temperature contours from above for nine x-direction and one y-direction fracture after 15.3
hrs.

For nine x-direction and one y-direction fracture, most flow is going through the middle x-direction
fracture and little fluid leaves through the y-direction fracture. This is because a pressure drop between
the injection and production well drives the flow mainly through the middle x-direction fracture.

Figure 6. Temperature contours from above for nine x-direction and two y-direction fractures after 11.06
hrs.

15
For nine x-direction and two y-direction fractures, some water leaves the middle x-direction fracture at
the first y-direction fracture and returns via the second y-direction fracture.

Figure 7. Temperature contours and flow vectors (kg/s per m2) from above for nine x-direction and nine
y-direction fractures after 112.8 hrs.

For nine x-direction and nine y-direction fractures, water is being spread throughout the whole reservoir.
Now also flow rate vectors (kg/s per m2) are shown. These indicate that water leaves the middle fracture
in first half and comes back in second half.

4.2 1x and 0y fractures


The simulation results for one x-direction and zero y-direction fractures are shown in this section,
complete results can be found in Appendix B. In table 6 the results are shown for 1 mm fracture aperture
and fracture permeability calculated by the different formulas k1, k2 and k3. Formula ,
was used to calculate the permeability of a fracture that was changed from aperture 3 mm to 1 mm due
to a confining pressure. For k0 the values of k1 and k2 were used.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k1 1 mm 33 2.045*107 No 0.76 days*
k2 1 mm 1.7 5.3*108 Yes 0.08 s
k3 using k0=k1 1 mm 33 2.0014*107 No 7.22 hrs*
k3 using k0=k2 1 mm 1.8 5.3*108 Yes 0.08 s

Table 6. Simulation results for 1 mm aperture with one x-direction and zero y-direction fractures.

16
From table 6 it is clear that the flow rate results differ substantially when using the different formulas k1
and k2, this is because the permeability for k2 was calculated to be lower than for k1. While for k1 the
maximum flow rate of 33 kg/s was reached, the flow rate at the production well was only 1.7 kg/s for k2.
Furthermore it is shown that the results for formula k3 depend strongly on the value used for k0 and give
similar results as when using formulas k1 or k2 alone.

Now the results obtained for 3 mm fracture aperture are shown in table 7 , for k3 the aperture was
changed from 1 mm to 3 mm due to the pumping pressure.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k1 3 mm 33 2.0017*107 No 2.19 days*
k2 3 mm 32.5 6.4*107 Yes 5.05 s
k3 using k0=k1 3 mm 33 2.0001*107 No 1.50 days*
k3 using k0=k2 3 mm 32.5 6.1*107 Yes 3.89 s

Table 7. Simulation results for 3 mm aperture with one x-direction and zero y-direction fractures.

Again flow rate results for k1 and k2 differ and flow rate results of k3 depend on the value used for k0.

Figure 8 shows the flow rate at the production well versus the fracture aperture, for the different
permeability formulas. It can be seen that the flow rate results for k1 and k2 differ substantially until they
approach the maximum flow rate of 33 kg/s. For k1 flow is still at its maximum for 300 µm while for k2 it
is already zero for 500 µm.

Fracture distribution 1x 0y
36
Flow rate production well (kg/s)

33
30
27
24
21 k1
18
15 k2
12 k3 k0=k1
9
6 k3 k0=k2
3
0
0 0.5 1 1.5 2 2.5 3 3.5

Fracture aperture (mm)

Figure 8. Flow rate at the production well versus fracture aperture, for different fracture permeability
formulas and fracture distribution 1x and 0y.

17
4.3 9x and 0y fractures
The results for nine x-direction and zero y-direction fractures are shown in table 8 and compared with
results for one x-direction and zero y-direction fractures.

Formulas Fractures Fracture Flow rate Pressure Simulation Simulation


used (number aperture production injection error time
and well(kg/s) well (Pa) (*=steady
orientation) state)
k1 1x and 0y 3 mm 33 2.0057*107 No 2.19 days*
k1 9x and 0y 3 mm 33 2.0017*107 No 0.63 days*
k2 1x and 0y 3 mm 32.5 6.4*107 Yes 5.05 s
k2 9x and 0y 3 mm 33 2.57*107 No 6.03 days*

Table 8. Simulation results for 1x 0y and 9x 0y fracture distributions.

Results show that when formula k2 is used to calculate the fracture permeability, the flow rate at the
production well is less for one x-direction and zero y-direction fractures than for nine x-direction and
zero y-direction fractures. This seems a bit strange since there is no flow possible between the parallel
fractures. This is caused by the fact that the fracture permeability was calculated to be smaller in the first
case because of the larger fracture spacing. The question now arises if it was right to take H0 to be the
fracture spacing. The formula for permeability was given by: . This was derived by equating the

flow rate through a fracture: , with the flow rate through a porous media of height H0:
. Now since the only porous material in this simulation is the fracture, it should be more
reasonable to set H0 equal to the fracture aperture h. In that case, k2 becomes: .
Alternatively only one material could be used, which means that one cell consists of a fracture in the
middle with non-permeable rock (anhydrite) around it and has height H0 equal to the fracture spacing
plus aperture (Figure 9). In that way the porosity of the material should be very low: , with ϕf
the porosity of the fracture alone.

Figure 9. Representation of a cell consisting of a single material.

18
4.4 9x and 1y fractures
The simulation results for nine x-direction and one y-direction fractures are shown in figure 10, complete
results can be found in Appendix C. Results for k1 and k2 differ substantially again.

Fracture distribution 9x 1y
36
Flow rate production well (kg/s)

33
30
27
24
21
18
15 k1
12
k2
9
6
3
0
0 0.5 1 1.5 2 2.5 3 3.5

Fracture aperture (mm)

Figure 10. Flow rate at the production well versus fracture aperture, for different fracture permeability
formulas and fracture distribution 9x and 1y.

4.5 9x and 2y fractures


The simulation results for nine x-direction and two y-direction fractures are shown in figure 11, complete
results can be found in Appendix D. Results show a similar pattern as for 9x and 1y fractures, with results
differing for k1 and k2.

Fracture distribution 9x 2y
36
Flow rate production well (kg/s)

33
30
27
24
21
18
15 k1
12
k2
9
6
3
0
0 0.5 1 1.5 2 2.5 3 3.5

Fracture aperture (mm)

Figure 11. Flow rate at the production well versus fracture aperture, for different fracture permeability
formulas and fracture distribution 9x and 2y.

19
4.6 9x and 9y fractures
The simulation results for nine x-direction and nine y-direction fractures are shown in figure 12,
complete results can be found in Appendix E. Again results for k1 and k2 differ and results for k3 depend
strongly on the value used for k0.

Fracture distribution 9x 9y
36
Flow rate production well (kg/s)

33
30
27
24
21 k1
18
15 k2
12
9 k3 k0=k1
6
3 k3 k0=k2
0
0 0.5 1 1.5 2 2.5 3 3.5

Fracture aperture (mm)

Figure 12. Flow rate at the production well versus fracture aperture, for different fracture permeability
formulas and fracture distribution 9x and 9y.

4.7 Overview different fracture distributions


In figure 13 flow rate-aperture results for different fracture distributions and permeability k2 is shown.
This shows that a high x-direction fracture density and connections between these fractures is required
to get a higher flow rate. It also shows that adding y-direction fractures lowers the flow rate, as more
fluid leaves the middle x-direction fracture.

Different fracture distributions


36
Flow rate production well (kg/s)

33
30
27
24
21
18 k2 1x0y
15
12 k2 9x9y
9
6 k2 1x9y
3
0
0 0.5 1 1.5 2 2.5 3 3.5

Fracture aperture (mm)

Figure 13. Flow rate at the production well versus fracture aperture, for different fracture distributions
and formula k2.

20
Figure 14 shows the pressure at the injection well versus the fracture aperture for different fracture
distributions and permeability k2. The results for fracture distribution 1x9y and permeability k2 can be
found in Appendix F. From figure 14 it can be seen that adding more fractures reduces the pressure at
the injection well, as more water can escape from the injection fracture. Furthermore it shows, that with
less fractures and smaller aperture, the pressure at the injection well goes up, in order to get a flow rate
of 33 kg/s at the production well. If this pressure exceeds above a certain limit, a simulation error will
occur. In reality such a high pressure would cause the rock to crack, however PetraSim ends the
simulation and gives an error. Also because the pressure goes up and the enthalpy of the injected water
remains constant, the temperature of the injected water will drop.

Different fracture distributions


8.00E+08
Pressure injection well (Pa)

7.00E+08
6.00E+08
5.00E+08
4.00E+08 k2 1x0y
3.00E+08 k2 9x9y
2.00E+08 k2 1x9y
1.00E+08
0.00E+00
0 0.5 1 1.5 2 2.5 3 3.5
Fracture aperture (mm)

Figure 14. Pressure at the injection well versus fracture aperture, for different fracture distributions and
formula k2

4.8 Different fracture porosity


Finally some simulations with different fracture porosity were carried out for the case of nine x-direction
and two y-direction fractures, with permeability calculated by k1 and fracture aperture 200 µm. The
results are shown in table 9.

Fracture porosity Flow rate Pressure injection Simulation error Simulation time
production well (Pa) (*=steady state)
well(kg/s)
0.9 31.5 6.4*107 Yes 0.41 s
0.5 (standard for 32 6.4*107 Yes 0.33 s
all simulations)
0.1 32.5 6.4*107 Yes 0.25 s

Table 9. Simulation results for different fracture porosities, for 200 µm aperture, formula k1 and fracture
distribution 9x and 2y.

21
No major differences were found for different fracture porosity, but for lower porosity the flow rate was
a little higher than for a high porosity.

5. Discussion
5.1 Flow behavior
In order to get flow going from an injection to a production well, the wells should be connected by
fractures. Adding more y-direction fractures lowers the flow rate through a single x-direction fracture
and reduces the pressure at the injection well, as more flow leaves the injection x-direction fracture.
However adding more y-direction fractures can also increase the flow rate at the production well
because it increases the connectivity of the x-direction fractures. When using multiple x-direction and y-
direction fractures, the fluid leaves the injection fracture in the first half length and returns in the second
half length. So fluid does not only take the shortest route but spreads throughout the reservoir, however
the largest flow rate is still via the shortest route. In general it can be said that more and better
connected x-direction fractures result in a higher flow rate at the production well.

5.2 Applicability of k1
Simulations showed that if formula was used to calculate the effective permeability of a single
fracture, the maximum flow rate of 33 kg/s at the production well was always reached for a single
fracture of 300 µm aperture and above. That means it is possible to let 33 kg/s water flow through a
single unconnected fracture of 300 µm aperture and width 10 m. To see if this is realistic, the right
numbers will be inserted into Darcy’s law: . In this case k=7.5*10-9 m2, A=0.03 m2, L=10 m,
Δp=3.8*107-2.0*107=1.8*107 Pa and µ=0.71*10-3 Pa*s (viscosity of water at 36°C, the mean fracture
temperature at the end of the simulation). This gives Q=0.056 m3/s=56 kg/s so it is possible. Further
simulations in fact showed that there was no error until the injection rate was 68 kg/s. The question now
is if the calculated permeability of 7500 Darcy is a realistic value. The Slochteren sandstone for example
has an XY-permeability of 250 mDarcy [Doddema, 2012], 30,000 times smaller. However in this
simulation, permeability was not averaged over the entire reservoir, but only the fracture material was
permeable. The maximum flow was also possible for 200 µm aperture, if the fracture spacing was equal
to or smaller than 1 m for x-direction fractures and 2 m for y-direction fractures. For 100 µm aperture
the maximum flow was not possible, even when 1 m fracture spacing was used in both x-direction and y-
direction direction. Formula k1 was derived under the assumption of describing a fracture by two smooth
parallel plates. Despite the fact that in reality fractures aren't smooth and parallel, formula k1 is still
useful when an effective aperture is used. Formula k1 can be used to calculate effective permeability of a
single fracture and is therefore useful in a discrete fracture network model, where the fractures are
explicitly defined.

22
5.3 Applicability of k2
Formula can only be used to calculate the effective permeability of a single fracture if H0 is

equal to the fracture aperture h. In that case, k2 becomes which is the same as formula k1.
Alternatively formula could be used to calculate the permeability of a single cell of the same material,
comprising a fracture surrounded by non-permeable rock with H0 equal to the fracture aperture h plus
the spacing between adjacent fractures. In that way the porosity of the material should be: ,
with ϕf the porosity of the fracture alone. However in this way flow is over the entire cell and not only
through the fracture, which makes it less realistic but easier and faster to develop for a large scale
reservoir model.

5.4 Applicability of k3
Formula , can be used to predict the fracture permeability after a stress induced
aperture change. However it requires an initial permeability k0, which should be calculated using formula
k1 or k2. It therefore strongly depends on these values, and simulation results differ only little from when
k1 or k2 is used to calculate permeability.

5.5 Recommendations
A smaller fracture spacing could be used to investigate the effect of increasing the fracture density even
more. Based on experimental results, Zulauf et al. (2011) found that for a fracture pattern in deformed
anhydrite, a smaller fracture aperture goes along with smaller fracture spacing [Zulauf et al., 2011] .
Therefore smaller fracture spacing should have been used for smaller fracture apertures. Wellbore data
is required to get an insight of the real fracture aperture and spacing, so that simulations can be more
specific for the situation. A uniform radial distribution of fractures could be simulated. This might be
useful because a radial fracture pattern might develop as a result of a salt layer below the anhydrite
pushing upwards. Furthermore scenarios where some of the fractures are not cutting through the whole
volume can be examined. Also a nonzero permeability and porosity for anhydrite can be used to make
the simulations more realistic. Different values for enthalpy of the injected water should be used if the
reservoir pressure goes up, so that the temperature of the injected water is always 10°C. Finally formula
k2 could be further examined, in the way that it calculates the permeability averaged over a single cell
comprising fracture and matrix. For instance this formula might not be useful for large fracture spacing,
but is rather useful for fractures with small fracture spacing.

23
6. Conclusion
Simulation results found by using formulas and for effective permeability of a single
fracture differ substantially. For the maximum flow rate of 33 kg/s requires an aperture of at least 200
µm, with 1 m x-direction and 2 m y-direction fracture spacing. For k2 the maximum flow rate of 33 kg/s
requires an aperture of at least 2 mm, with 1 m x-direction and 10 m y-direction fracture spacing.
Differences in results can be explained by the fact that k2 can only be used to calculate the permeability
of a single fracture if H0 is set to h, in which case k2 becomes the same as k1. Formula k1 is applicable in
discrete fracture network models, where fractures are explicitly defined. Formula k2 can be used to
calculate the permeability of an entire cell consisting of a fracture and its impermeable surrounding. For
this cell, height H0 is equal to the fracture aperture plus the fracture spacing. This is easier and faster to
use in a large scale reservoir model, as the fractures do not have to be defined individually but rather
one single material can be used. The disadvantage of this, is that it is less realistic than when using two
separate materials for fracture and matrix, because in this case the permeability is the same over the
entire cell including the non-fracture part. Formula can be used to predict the fracture
permeability after a stress induced aperture change. However this formula depends strongly on the
initial permeability value k0, and therefore on the formula used to calculate this value

24
7. Acknowledgements
I want to thank prof. dr. Rien Herber for being my first supervisor and providing me a place at the Geo-
energy group. I also want to thank him for his advice and support during the project. I want to thank Alex
Daniilidis in particular, for being my daily supervisor during this project. His support and help made it
possible to finish this thesis. Finally I want to thank prof. dr. Harro Meijer for being the second supervisor
of this project.

25
8. References

Carman, P. C. (1956): Flow of Gases through Porous Media. Academic Press, New York.

Christiansen, R. L. and S. M. Howarth (1995):


Literature review and recommendation of methods for measuring relative permeability of anhydrite from
the Salado Formation at the Waste Isolation Pilot Plant. Sandia National Labs, Albuquerque, NM (United
States).

Doddema, L. (2012), The influence of reservoir heterogeneities on geothermal doublet performance,


RUG, Groningen, The Netherlands.

Gilman, J. R. (2003): Practical Aspects of Simulation of Fractured Reservoirs; iReservoir.com, Baden-


Baden, Germany.

Pruess, K. (1983): GMINC - A Mesh Generator for Flow Simulations in Fractured Reservoirs; Earth Sciences
Division, Lawrence Berkeley National Laboratory, Berkeley CA USA.

Pruess, K., C. Oldenburg, and G. Moridis (Eds.) (1999): TOUGH2 User's Guide, Version 2.0; Earth Sciences
Division, Lawrence Berkeley National Laboratory University of California, Berkeley, USA.

Raven, K. G. and J. E. Gale (1985): Water flow in a natural fracture as a function of stress and sample
size. Int. J. Rock Mech. Min. Sci. Geomech. 22 251-261.

Roman, A., G. Ahmadi, K. A. Issen, and D. H. Smith (2012): Permeability of fractured media under
confining pressure: A simplified model; Open Pet.Eng.J.Open Petroleum Engineering Journal 5 36-41.

Sarkar, S., M. N. Toksöz, and D. R. Burns (2004): Fluid Flow Modeling in Fractures; Massachusetts
Institute of Technology, Cambridge, USA.

Thunderhead Engineering (2012): PetraSim5 User Manual; Thunderhead Engineering, Manhattan, USA.

Warren, J. E. and P. J. Root (1963): The Behavior of Naturally Fractured Reservoirs; Society of Petroleum
Engineers Journal 228 245-246-255.

Wilkes, J. O. (1999): Fluid Mechanics for Chemical Engineers; Prentice Hall.

Witherspoon, P. A., J. S. Y. Wang, K. Iwai, and J. E. Gale (1980): Validity of cubic law for fluid flow in a
deformable rock fracture. Water Resour. Res 16 1016-1024.

Zhang, J., W. B. Standifird, J. C. Roegiers, and Y. Zhang (2007): Stress-Dependent Fluid Flow and
Permeability in Fractured Media: from Lab Experiments to Engineering Applications; Rock Mech Rock Eng
40 3-21.

Zimmerman, R. W. and G. S. Bodvarsson (1995): Hydraulic Conductivity of Rock Fractures ; Transport in


Porous Media 23 1-30.

26
Zulauf, J., G. Zulauf, J. Hammer, and F. Zanella (2011): Tablet boudinage of an anhydrite layer in rock-salt
matrix: Results from thermomechanical experiments; Journal of Structural Geology 33 1801-1815.

27
9. Appendix A: Fracture permeability and productivity index calculated
by the different formulas

Fracture aperture Permeability(m2) Productivity


index(m3)
3 mm 7.5*10-7 9.4*10-5
2 mm 3.3*10-7 4.1*10-5
1 mm 8.33*10-8 1.05*10-5
500 µm 2.08*10-8 2.62*10-6
300 µm 7.5*10-9 9.42*10-7
200 µm 3.33*10-9 4.18*10-7
100 µm 8.33*10-10 1.05*10-7

Table 10. Fracture permeability and productivity index for various fracture apertures calculated by using
formula k1.

Fracture X-direction Y-direction X-direction Y-direction X-direction Y-direction PI(m3)


aperture fractures fractures fracture fracture fracture fracture
spacing(m) spacing(m) permeabiltiy(m2) permeability(m2)
3 mm 1 0 10 X 2.3*10-10 x 2.9*10-8
3 mm 9 0 1 X 2.3*10-9 x 2.9*10-7
3 mm 9 1 1 10 2.3*10-9 2.3*10-10 2.9*10-7
3 mm 9 2 1 4 2.3*10-9 5.6*10-10 2.9*10-7
3 mm 9 9 1 1 2.3*10-9 2.3*10-9 2.9*10-7
3 mm 1 9 10 1 2.3*10-10 2.3*10-9 2.9*10-8
2 mm 1 0 10 X 6.67*10-11 x 8.38*10-9
2 mm 1 1 10 10 6.67*10-11 6.67*10-11 8.38*10-9
2 mm 1 9 10 1 6.67*10-11 6.67*10-10 8.38*10-9
2 mm 5 1 2 10 3.33*10-10 6.67*10-11 4.18*10-8
2 mm 9 1 1 10 6.67*10-10 6.67*10-11 8.38*10-8
2 mm 9 2 1 4 6.67*10-10 1.67*10-10 8.38*10-8
2 mm 9 5 1 2 6.67*10-10 3.33*10-10 8.38*10-8
2 mm 9 9 1 1 6.67*10-10 6.67*10-10 8.38*10-8
1 mm 1 0 10 X 8.33*10-12 x 1.05*10-9
1 mm 1 1 10 10 8.33*10-12 8.33*10-12 1.05*10-9
1 mm 1 9 10 1 8.33*10-12 8.33*10-11 1.05*10-9
1 mm 5 1 2 10 4.16*10-11 8.33*10-12 5.24*10-9
1 mm 9 1 1 10 8.33*10-11 8.33*10-12 1.05*10-8
1 mm 9 2 1 4 8.33*10-11 2.08*10-11 1.05*10-8
1 mm 9 5 1 2 8.33*10-11 4.16*10-11 1.05*10-8
1 mm 9 9 1 1 8.33*10-11 8.33*10-11 1.05*10-8
500 µm 1 0 10 x 1.04*10-12 x 1.31*10-10
500 µm 1 1 10 10 1.04*10-12 1.04*10-12 1.31*10-10
500 µm 5 1 2 10 5.21*10-12 1.04*10-12 5.24*10-9
500 µm 9 1 1 10 1.04*10-11 1.04*10-12 1.05*10-9

28
500 µm 9 2 1 4 1.04*10-11 2.60*10-12 1.05*10-9
500 µm 9 5 1 2 1.04*10-11 5.21*10-12 1.05*10-9
500 µm 9 9 1 1 1.04*10-11 1.04*10-11 1.05*10-9
100 µm 1 0 10 x 8.33*10-15 x 1.05*10-12
100 µm 9 1 1 10 8.33*10-14 8.33*10-15 1.05*10-11
100 µm 9 2 1 4 8.33*10-14 2.08*10-14 1.05*10-11
100 µm 9 9 1 1 8.33*10-14 8.33*10-14 1.05*10-11
100 µm 1 9 10 1 8.33*10-15 8.33*10-14 1.05*10-12

Table 11. Fracture permeability and productivity index for various fracture apertures calculated by using
formula k2.

Aperture change Permeability(m2) Productivity


index(m3)
From 1 mm to 3 mm 2.25*10-6 2.82*10-4
From 3 mm to 1 mm 2.78*10-8 3.49*10-6

Table 12. Fracture permeability and productivity index after an stress induced aperture change
calculated by using formula k3 and k0=k1.

Aperture Fractures (number and X-direction Y-direction X-direction Y-direction PI(m3)


change orientation) fracture fracture fracture fracture
spacing(m) spacing(m) permeabilitiy(m2) permeability(m2)
From 1 mm 1x and 0y 10 x 2.25*10-10 x 2.83*10-8
to 3 mm
From 3 mm 1x and 0y 10 x 8.5*10-12 x 1.1*10-9
to 1 mm
From 1 mm 9x and 9y 1 1 2.25*10-9 2.25*10-9 2.83*10-7
to 3 mm
From 3 mm 9x and 9y 1 1 8.5*10-11 8.5*10-11 1.1*10-8
to 1 mm

Table 13. Fracture permeability and productivity index after an stress induced aperture change
calculated by using formula k3 and k0=k2.

29
10. Appendix B: 1x and 0y simulation results

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k1 3 mm 33 2.0057*107 No 2.19 days*
k1 2 mm 33 2.0057*107 No 2.19 days*
k1 1 mm 33 2.045*107 No 1.92 days*
k1 500 µm 33 2.37*107 No 1.25 days*
k1 300 µm 33 3.8*107 No 1.47 hrs*
k1 200 µm 32.5 6.4*107 Yes 0.32 s
k1 100 µm 32.2 2.2*108 Yes 0.03 s

Table 14. Simulation results for formula k1 with one x-direction and zero y-direction fractures.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
K2 3 mm 32.5 6.4*107 Yes 5.05 s
K2 2 mm 32.3 1.54*108 Yes 0.93 s
K2 1 mm 1.7 5.3*108 Yes 0.08 s
K2 500 µm 0 7.0*108 Yes 0.02 s
K2 100 µm 0 7.56*108 Yes 0.00 s

Table 15. Simulation results for formula k2 with one x-direction and zero y-direction fractures.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k3 using k0=k1 3 mm 33 2.0001*107 No 1.50 days*
k3 using k0=k1 1 mm 33 2.0014*107 No 7.22 hrs*
k3 using k0=k2 3 mm 32.5 6.1*107 Yes 3.89 s
k3 using k0=k2 1 mm 1.8 5.3*108 Yes 0.08 s

Table 16. Simulation results for formula k3 with one x-direction and zero y-direction fractures.

30
11. Appendix C: 9x and 1y simulation results

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k1 3 mm 33 2.0017*107 No 0.64 days*
k1 300 µm 33 3.8*107 No 1.43 hrs*
k1 200 µm 32.5 6.4*107 Yes 0.33 s
k1 100 µm 18 1.9*108 Yes 0.04 s

Table 17. Simulation results for formula k1 with nine x-direction and one y-direction fractures.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
K2 3 mm 33 2.57*107 No 6.81 days*
K2 2 mm 33 5.2*107 No 1.53 days*
K2 1 mm 21.5 1.95*108 Yes 0.35 s
K2 500 µm 0.24 5.6*108 Yes 0.03 s
K2 100 µm 0 7.5*108 Yes 0.00 s

Table 18. Simulation results for formula k2 with nine x-direction and one y-direction fractures.

31
12. Appendix D: 9x and 2y simulation results

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k1 3 mm 33 2.0012*107 No 11.06 hrs*
k1 300 µm 33 2.5*107 No 5.43 hrs*
k1 200 µm 32 6.4*107 Yes 0.33 s
k1 100 µm 26 1.42*108 Yes 0.06 s

Table 19. Simulation results for formula k1 with nine x-direction and two y-direction fractures.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time
well(kg/s) (Pa)
K2 3 mm 33 2.43*107 No 2.67 days*
K2 2 mm 33 4.3*107 No 1.08 days*
K2 1 mm 20 1.65*108 Yes 0.42 s
K2 500 µm 0.13 5.5*108 Yes 0.03 s
K2 100 µm 0 7.5*108 Yes 0.00 s

Table 20. Simulation results for formula k2 with nine x-direction and two y-direction fractures.

32
13. Appendix E: 9x and 9y simulation results

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k1 3 mm 33 2.0005*107 No 0.53 days*
k1 2 mm 33 2.85*107 No 1.08 days*
k1 200 µm 33 3.7*107 No 0.57 hrs*
k1 100 µm 19 9.7*107 Yes 0.08 s

Table 21. Simulation results for formula k1 with nine x-direction and nine y-direction fractures.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
K2 3 mm 33 2.17*107 No 2.13 days*
K2 2 mm 33 2.85*107 No 1.08 days*
K2 1 mm 29 9.75*107 Yes 0.69 s
K2 500 µm 0.04 4.1*108 Yes 0.06 s
K2 100 µm 0 7.5*108 Yes 0.01 s

Table 22. Simulation results for formula k2 with nine x-direction and nine y-direction fractures.

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time (*=steady
well(kg/s) (Pa) state)
k3 using k0=k1 3 mm 33 2.00016*107 No 5.54 hrs*
k3 using k0=k1 1 mm 33 2.04*107 No 10.76 hrs*
k3 using k0=k2 3 mm 33 2.17*107 No 1.91 days*
k3 using k0=k2 1 mm 30 9.5*107 Yes 0.79 s

Table 23. Simulation results for formula k3 with nine x-direction and nine y-direction fractures.

33
14. Appendix F: 1x and 9y simulation results

1x and 9y fractures

Formulas used Fracture Flow rate Pressure Simulation Simulation


aperture production injection well error time(*=steady
well(kg/s) (Pa) state)
K2 3 mm 33 5.4*107 No 3.32 days*
K2 2 mm 25 1.38*108 Yes 1.07 s
K2 1 mm 0.022 3.6*108 Yes 0.15 s
K2 100 µm 0 7.6*108 Yes 0.00 s

Table 24. Simulation results for formula k3 with one x-direction and nine y-direction fractures.

34

You might also like