The Development, Validation and Application of PRSS Novel Upscaling Technology

1
The Development, Validation and Application of
PRSS Novel Upscaling Technology
Dr. Nasir Hj. Darman
2
Presentation Outline

• Background
• The New Optimal Grid Coarsening Scheme
• The New Pseudo Generation Method
• Application of the New Technologies
• Summary and Conclusions

3
Presentation Outline

• Background
• The New Optimal Grid Coarsening Scheme
• The New Pseudo Generation Method
• Application of the New Technologies
• Summary and Conclusions

4
Background

• Upscaling What is it ?

Geological model
Simulation model
Small scale (Millions of grid blocks)
Big scale (Thousands of grid blocks)
Upscaling
To decrease the resolution of the fine grid
models (to an acceptable level), without losing
the accuracy of its calculation
5
Background

• Why do we need such a fine resolution ?

Heterogeneity occurs at all scale, even at small
scale!
Capturing this heterogeneity (at all scale) is
the key to build a good simulation model
6
Background

• Upscaling Why do we need it ?

Geological model
Simulation model
Increase in calculation difficulty
Initial time only (in-place estimation)
Currently available computer systems are not
capable of handling direct calculations of
geological models in the simulation mode.
Upscaling process is needed!
Simulation through time (reserves estimation)
7
Background

• Upscaling How do we approach the problem ?

Upscaling
Pseudo Functions
Optimal Grid coarsening
averaged saturation equations

• To correct for numerical dispersion
• To capture the interaction of fluid flow and
  fine scale heterogeneity

• To intelligently coarsen the coarse grid models
  so that upscaling error can be minimized

8
Presentation Outline

• Background
• The New Optimal Grid Coarsening Scheme
• The New Pseudo Generation Method
• Application of the New Technologies
• Summary and Conclusions

9
The New Optimal Grid Coarsening Scheme

• We have derived a new averaged saturation
  equation

• Has been presented at the 8th European Conference
  on the Mathematics of Oil Recovery (September
  2002)
• Published in the SPEJ (without capillary
  pressure)

10
The New Optimal Grid Coarsening Scheme
The equation describes the changes in fluid
saturation at the coarse-scale model
Can be used to determine the coarse grid
configuration
The basic concept is to reduce variability of
certain fine grid fluctuating properties by
grouping them together in the coarse grid model
11
The New Optimal Grid Coarsening Scheme
Which fluctuating higher moment to use is
dependent on the balance of fluid forces.
and
Capillary Dominated
!
Gravity Dominated
Viscous Dominated
and
12
The New Optimal Grid Coarsening Scheme
We have validated the technique using many test
cases and production scenarios
One of the test cases ..
(random permeability distribution. 80-layer fine
grid)
13
The New Optimal Grid Coarsening Scheme
Procedures
start
Run the fine grid (true answer)
Calculate rms error (difference between fine
coarse grid)
Visually determine which moments can be used to
minimize error
Based on the fine grid run, determine the optimal
coarse grid configuration for 40,30,20,10 and 1
layer (minimization of the moment)
Plot the rms error versus number of coarse grid
layer
stop
Run the coarse grid
14
The New Optimal Grid Coarsening Scheme
Case where Gravity and Viscous are important
From the equation, we know that SS is the best
variable to be used
40-layer coarse grid
Fine grid (true answer)
SS
S.dS/dx and kx.dS/dx fail to minimize the
resulting error satisfactorily
15
The New Optimal Grid Coarsening Scheme

A

• The strength of the new method is that it can
  predict the resultant error given a required
  coarse grid layer requirement
• Users can include their technical judgement in
  balancing between practicality and accuracy

8.00E-02
7.00E-02
B
6.00E-02
5.00E-02
RMS error
4.00E-02
C
3.00E-02
2.00E-02
1.00E-02
0.00E00
0
5
10
15
20
25
30
35
40
No. of coarse grid layer
ltS'S'gt
ltS'.dS/dxgt
ltkx'.dS/dxgt
16
Presentation Outline

• Background
• The New Optimal Grid Coarsening Scheme
• The New Pseudo Generation Method
• Application of the new Technologies
• Summary and conclusions

17
The New Pseudo Generation Method

• Currently available pseudo generation methods
  such as Kyte Berry or Stone have limited
  capabilities
• We have derived a new pseudo generation
  methodcalled the TW method (transmissibility
  weighted)
• Refer to SPE 51941 (also has been published in
  the Computational Geosciences special issue on
  Upscaling)
• Believe to work better than currently available
  pseudo techniques (in all environments)

18
The New Pseudo Generation Method
We have validated the technique using many test
cases and production scenarios
Some of the test cases ..
(fine grid size 103 x 1 x 40 0.2 cm in each
directions)
19
The New Pseudo Generation Method
We have validated the technique using many test
cases and production scenarios
Some of the test results ..
20
The New Pseudo Generation Method
We have validated the technique using many test
cases and production scenarios
Some of the test results ..
21
The New Pseudo Generation Method
We have validated the technique using many test
cases and production scenarios

• In all of our test cases and simulated production
  scenarios, the TW method performs better or equal
  to other currently available technique.
• The other strength of the TW method is that we
  develop the method systematically in a controlled
  environments

We know its strength and weakness relative to
other pseudo generation methods. Refer to SPE
66377
22
Presentation Outline

• Background
• The New Optimal Grid Coarsening Scheme
• The New Pseudo Generation Method
• Application of the new Technologies
• Summary and conclusions

23
Application of the New Technologies
Reservoir X
Fault 1
Growth Fault

• Simple roll-over anticline.
• Based on seismic, the reservoirs were divided
  into the shallow marine environment in the east
  and the incised valley channel-dominated
  reservoirs in the west.
• Production started in 1975
• 90 (300 ft) x 62 (200 ft) x 119 (3 ft)
• Total of 185,323 active fine grids

Fault 2
24
Application of the New Technologies
Reservoir X

• Not very heterogenous
• Good testing model
• Can focus on the physics rather than other
  phenomenon
• Allow identification of the weakness easily
• Allow corrective measure to be done easily

Cross-section diagram showing porosity attributes
25
Application of the New Technologies

• Upscaling
• Two 2-D cross-sectional fine grid models were
  developed to come out with the best coarsening
  schemes.
• SW-NE direction
• SE-NW direction

26
Application of the New Technologies
Model 1 (74 x 1 x 119)
27
Application of the New Technologies

• Based on the saturation trend, we group the fine
  grid layers
• Reduce to 16 layers

28
Application of the New Technologies
Fine Grid vs. Coarse Grid Results (W-3)
29
Application of the New Technologies
Fine Grid vs. Coarse Grid Results (W-29)
Water-cut
Pressure
Coarse grid
Fine grid
Oil Rate
GOR
30
Application of the New Technologies
Fine Grid vs. Coarse Grid Results (W-1)
Water-cut
Pressure
Coarse grid
Fine grid
Oil Rate
GOR
31
Application of the New Technologies
Model 2 (22 x 1 x 119)
32
Application of the New Technologies

• Apply the same coarsening strategy as Model 1
• (Dont calculate)

33
Application of the New Technologies
Fine Grid vs. Coarse Grid Results (W-1)
Water-cut
Pressure
Coarse grid
Fine grid
Oil Rate
GOR
34
Application of the New Technologies
Fine Grid vs. Coarse Grid Results (W-11)
Water-cut
Pressure
Coarse grid
Fine grid
Oil Rate
GOR
35
Application of the New Technologies
MODEL COMPARISON Net Pore Volume
36
Application of the New Technologies
Manage to reduce 119 layers fine grid model to
only 16 layers without losing much accuracy
Reduce running time
Reduce cost
Increase profitability
37
Presentation Outline

• Background
• The New Optimal Grid Coarsening Scheme
• The New Pseudo Generation Method
• Application of the new Technologies
• Summary and conclusions

38
Summary and Conclusions

• PRSS has successfully developed a new pseudo
  generation method and optimal grid coarsening
  strategy to improve the whole upscaling procedures

Derivation
Validation
Application
(preliminary)

• Currently in the stage of packaging the two
  techniques (as well as existing available
  techniques) into a user-friendly software

Derivation
Validation
Application
Software
(preliminary)
39
Acknowledgements

• PETRONAS MANAGEMENT
• PETRONAS CARIGALI (PCSB)
• PRSS UPSCALING PROJECT TEAM
• En. Shahberi Yusof (PCSB) and En. A Nazhri M Zain
  (PRSS)

40
Thank you