RESERVOIR MODELING AND
SIMULATION
ECLIPSE PROJECT 1 (Marks: 10%)
Name: Nabila Ghaisani
Student ID: 22309
Signature: __________________
Result (DATA for Project 2)
Part 2.A
The keyword FWIT is Field Water Injection
FWCT & FOE vs FWIT Total. It shows how much water is injected in to
1 the reservoir.
0.8 Water cut is defined by Schlumberger as The
0.6 ratio of water produced compared to the volume
0.4 of total liquids produced. This term will be related
0.2 to keyword FWCT.
0
0 10000000 20000000 30000000 40000000
While Oil Recovery is the ratio of oil recovered
from the reservoir compared to the initial oil in
FWCT FOE the reservoir. To be noted, FOE is accumulative
term and therefore It will keep increasing. This
Figure 1 : FWCT & FOE vs FWIT for part A term will be related to keyword FOE.
We can see when there is no water injected, the FWCT is zero due to no water produced yet and same goes
for FOE. As the simulation started, the FOE will start increase, shows that oil has been produced from the
reservoir while the FWCT is still zero. At approximately 1000000 stb water injected, finally the FWCT
starts to increase. This shows that the water is started to produce along the water.
The last FOE recorded is at value 0.652147, we can conclude that 65% of the oil in the reservoir is
recovered. The last FWCT recorded is 0.951392, that shows that the percentage of water produced at that
time is 95% total of liquid produced, leaving 5% of oil.
Part 2.B
Modification :
FOE & FWCT vs FWIT
- Putting the highest permeability in the
1
last 5 layers (Layer 11 to 15)
0.8 - Alter the PERMZ, PORO, NTG and
SATNUM keywords to reflect the layer
0.6
changes.
0.4
0.2
0
0 10000000 20000000 30000000
FOE - 2A FWCT - 2A
FOE - 2B FWCT - 2B
Figure 2 : FOE & FWCT vs FWIT for part B
Part A
Timestep 2 Final Timestep
Part B
Timestep 2 Final Timestep
Figure 3 : Grid Saturation Profile for Part A & B
It can be seen from timestep 2 that the oil saturation for each part is concentrated in different layers ( For
Part A, in the middle and part B in the bottom). From the final timestep, we can see that the water displaces
more oil in part A rather than part B. It is because the oil will flow faster in part A in the middle layers and
helped by the gravity, the water will be pulled down and replacing the oil in the bottom layers also.
However, in part B the bottom layers have the highest permeability and helped by the gravity, the water
will displace the oil faster only in bottom layers while the upper part will have the water displacing the oil
much slower. Therefore, it shows in the graph above the FOE in part B ( 0.55842358) is lower than part A
(0.652147) .
Part c
The datafile for Part 2C would then be used for
base file for the following parts of the project. The
main difference between part c and previous ones
in the placement of the highest permeability (100
md)
Part 2.D
Modification on part D :
- Changing injection rate from 11,000 stb water/day only 1,100 stb/day
- Changing production rate from 10,000 stbl/day only 1,000 stbl/day
- Increasing timestep from 285 to 2850 days each.
Analysis
Part A- D has different properties compared to each other such as:
1. Part A : Highest permeability = Layer 6-10
2. Part B : Highest Permeability = Layer 11 15
3. Part C : Highest Permeability = Layer 1- 5
4. Part D : Similar grid properties but different timestep.
In comparison, the oil will flow the fastest in the layer with higher [Link] part D, compared
to other part have different result where the oil displaced more in the bottom layer when the highest
permeability is in the first 5 layers. This is caused due to 2 factor. First is the low rate of injection and
production, the other reason is the long timestep. To be noted that timestep 2 in part D is 5700 days while
in other parts its 575 days. Added with the low rates, the gravity will influence the displacement more than
the q. Therefore it shows how the displacement is seen more in the bottom layer rather than top layers.
Same goes for timestep 5 where the oil is displaced more in the layers with higher permeability. Again, the
gravity will influence more than the rates of injection. The displacement in part D then will be more even
compared with other parts. However, because of the long timestep and low rate the FOE and FWCT of part
D will reflect part C because of the same grid properties although its very slow compared with part C.
Timestep 2
Part A Part B
Part C Part D
Figure 4 : Grid Saturation Profile on Timestep 2 for part A-D
Timestep 5
Part A Part B
Part C Part D
Figure 5 : Grid Saturation Profile on Timestep 5 for part A-D
FOE vs FWIT ( Part a-d)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 5000000 10000000 15000000 20000000 25000000 30000000 35000000
FOE-A FOE-B FOE-C FOE-D
Figure 6 : FOE vs FWIT for Part A-D
FWCT vs FWIT (part a-d)
1
0.8
0.6
0.4
0.2
0
0 5000000 10000000 15000000 20000000 25000000 30000000 35000000
FWCT - A FWCT-B FWCT-C FWCT-D
Figure 7 : FWCT vs FWIT for Part A-D
Part 2.E
Modification : changing y-grid value
Timestep 2 Timestep 5
Figure 8 : Grid Saturation Profile on Part E
Analysis :
Here in part E, we made the reservoir grids with variable dy. This will result in reservoir that will be
imitating radial reservoir with the nearest dy with small value and getting bigger with further grid. This
arrangement will of course affect the flow distribution of liquid. You can see that compared with former
parts, this part is tend to have the displacement not clearly noticed. This is not because of the slow rate of
the displacement, but this is more on the even distribution. The basic of the flow is to have pressure
difference, and from high pressure to low pressure. It will not see the direction and therefore the
displacement will try to even the distribution. With bigger dy as the water displace the oil, from this
perspective it will only show a slow displacement.
Part 2.F
Modification : changing the anisotrophy (kv/kh) to 1
Timestep 2 Timestep 5
Figure 9 : Grid Saturation Profile on Part F
Part 2.G
Modification :
- Insert keyword MULTZ in the GRID section as transmissibility barrier between each layer
FOE vs FWIT (part c,e,f,g)
0.8
0.6
0.4
0.2
0
0 500 1000 1500 2000 2500 3000
FOE - C FOE-E FOE-F FOE-G
Figure 10 : FOE vs FWIT for Part C,E,F,G
FWCT vs FWIT (part c,e,f,g)
1.2
0.8
0.6
0.4
0.2
0
0 500 1000 1500 2000 2500 3000
FWCT - C FWCT - E FWCT - F FWCT - G
Figure 11 : FWCT vs FWIT for Part C,E,F,G
Timestep 2
Part C Part E
Part F Part G
Figure 12 : Grid Saturation Profile on Timestep 2 for Part C,E,F,G
Timestep 5
Part C Part E
Part F Part G
Figure 13 : Grid Saturation Profile on Timestep 5 for Part C,E,F,G
In reality most reservoirs are not homogeneous in terms of permeability (isotropic permeability). Usually
permeability varies significantly between the vertical and horizontal planes within a formation with normal
value is preset already (kv = 0.1 kh).
Part E shows slow displacement (explanation already exists in previous part) of oil and therefore the FWCT
is almost zero, and the FOE is much lower compared with other parts.
Here in part f, the difference with part c as the base case is the anisotrophy which is the ratio of vertical
permeability to its horizontal permeability 1 instead of 0.1. By having the vertical permeability same with
the horizontal permeability, it will be easier for the gravity to pull down the water injected. As the flows
will be faster in the first 5 layers, the water would be pulled down by the gravity making the distribution
more even. This will affect the FOE & FWCT of part F as the oil recovered (75% of total oil) is much
higher in amount due to even distribution and water already reached the production well faster compared
with other parts.
However in part G, the transmissibility barrier will cause the gravity not able to drive the water down to
bottom layers. So the flow force will be the only factor that controls the distrtibution. The water will displace
oil differently in each layer. The first 5 layer with the higher permeability will have faster displacement and
the rest will have slower displacement. This will cause low FOE & FWCT with oil recovery 55% but not
as low as FOE of Part E.
Part 2.H
Modification = setting Pc to 0 (zero)
Timestep 2 Timestep 5
Figure 14 : Grid Saturation Profile on Part H
FOE vs Time ( Part C & H)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 500 1000 1500 2000 2500 3000
C H
Figure 15 : FOE vs Time for Part C & H
Capillary pressure affects waterflood performance and engineering calculations because the extent to which
the water/oil flood front is vertically and horizontally "smeared out" during the waterflood is controlled by
the Pc/Sw imbibition curve.
Capillary pressure tends to favor imbibition of the wetting phase into a cell or the cells retention of the
wetting phase. Therefore, Pcwo should increase oil recovery in the water-oil displacement because water will
imbibe into the the tight layer from its adjacent permeable layer, causing oil to flow out of the tight layer
by counter-current imbibition.
However by having Pc=0, it means that the recovery of the oil is reduced. It can be seen from the graph
where the FOE of part H is slightly lower than part C
Part 2.I
Modification :
- coarsening the grid model
FOE vs Time ( Part C & Coarsening)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 500 1000 1500 2000 2500 3000
C H
Figure 16: FOE vs Time for Part C & C-Coarsening
Part 2.J
Modification:
-refine the grid model by adding autoref
FOE vs Time (2C & 2C refine)
0.8
0.6
0.4
0.2
0
0 500 1000 1500 2000 2500 3000
REFINE PART C
Figure 17: FOE vs Time for Part C & C-Refine
FWCT vs Time (2C & 2CREFINE)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 500 1000 1500 2000 2500 3000
FWCT - 2c FWCT - 2CREFINE
Figure 18: FWCT vs Time for Part C & C-Refine
In many problems we need a higher resolution (finer grid) than our grid permits. We can give a good
estimate for time and position of the breakthrough in the well. Also, the cells are sufficiently small that they
can be classified as either water filled or oil filled. It can also give more accurate estimate of the
displacement including capillary pressure (the value is almost similar to the base case) . Capillary pressure
will definitely affect this model. When the same problem is modelled on a coarser grid, the details of
previous model will be not clearly defined & it goes for capillary pressure as well showing large difference
with the base case. This means that the capillary pressure will not that matter in the coarse model compared
with refined model.
Using finer grid on the entire reservoir is typically not possible due to memory limitations and computing
time. One possibility is to extend the fine grid in all directions with coarser cells, coarsening the grid. This
is, however, not a recommended solution, since the resulting long and narrow cells are sources of
computational errors, especially when the size difference between large and small cells in the grid becomes
too large. Grid coarsening may be used to amalgamate cells in regions of the reservoir which are less
important to speed up simulation run times.
FOE vs Time ( Part Refine & Coarsening)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 500 1000 1500 2000 2500 3000
REFINE H
Figure 19 : FOE vs Time for Part C-Refine & C-Coarsening
Sensitivities:
In the technology, surfactant polymer is injected to the reservoir to reduce an interfacial tension between
oil and water and is able to wipe out the trapped oil from the reservoir rock and hence increase the oil
production. The primary purpose of adding polymer to most polymer waterfloods is to increase the viscosity
of the flood water.
Figure 20 : FOE vs Time for Part C, Viscous Oil & Polymer
Part C (Base Case)
4500 800 1.1
=
6000 800 2.0 1.1
= 1.7404 cp
kro krw oil water Mobility (M)
0.9 0 0.5171
0.3 0.2 0.4224
0.1 0.4 1.7404 0.8 0.5575
0 0.55 0.6875
0 1 1.25
Sensitivities (Viscous Oil)
4500 800 5.5
=
6000 800 10 5.5
= 8.7019 cp
Mobility
kro Krw oil water
(M)
0.9 0 0.103425689
0.3 0.2 0.2845
0.1 0.4 8.7019 0.8 0.5114
0 0.55 0.6875
0 1 1.25
Sensitivities (Polymer)
Water viscosity = 10 cp
Mobility
kro krw oil water
(M)
0.9 0 0.1034
0.3 0.2 0.0544
0.1 0.4 8.7019 10 0.0514
0 0.55 0.055
0 1 0.1
Mobility ratio is the mobility of an injectant divided by that of the fluid it is displacing, such as oil.
In the base case, the oil recovery is quite high because the value of viscosity of oil and water are deemed
quite low. If we change the viscosity of the oil without changing the water, that will result in slow rate of
production as in Darcys law:
If the viscosity of oil decrease then the rate will increase and vice versa.
Now if we modify the viscosity of the water as well, we would then get the similar FOE with the base
case with the difference in the mobility. Therefore we can conclude that adding polymer improve the
sweep efficiency and the recovery in the viscous oil.