REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy •...
Transcript of REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy •...
![Page 1: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/1.jpg)
SMART CFD PROXYREPLICATING CFD RESULTS ACCURATELY & AT HIGH SPEED
Shahab D. Mohaghegh1, Mehrdad Shahnam2, Amir Ansari1, Ali Takbiri1, Ebrahim Fathi1, and Jean F. Dietiker2
August 2016
1West Virginia University2DOE-NETL
![Page 2: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/2.jpg)
Outline
• Smart Proxy• Engineering Application of Big Data Analytics
• Original Application in Reservoir Modeling• Tracking Pressure & Saturation Changes at the Grid Block Level
• Application to CFD• Multi-Phase Fluidized Bed (Proof of Concept)
![Page 3: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/3.jpg)
Computational Science
• Numerical solution to complex, multi-physic, non-linear, partial differential equation emerged, when
• Engineering models became too complex to be solved analytically, • High speed computers became widely available.
![Page 4: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/4.jpg)
Computational Footprint
• Our models soon increased in size such that use of super computers and CPU clusters became a necessity to solve serious problems.
• Proxy models are used to address the extensive computational cost of the numerical simulation models.
![Page 5: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/5.jpg)
Proxy Models
• Proxy models attempt to behave like the original models but with lower computational cost.
• The current paradigm for developing proxy models is to simplify the original model.
• Existing proxy models are either:• Statistics-based (response surfaces – curve fit at specific locations)• Physics-based (Reduced Physics)• Coarse Models (low resolution)
![Page 6: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/6.jpg)
Paradigm Shift in Proxy Modeling
• In this new paradigm of developing proxy models, • The proxy model represents the entire system.• Physics of the model is NOT simplified.• Model resolution is NOT reduced.
• In this new paradigm, data (generated by CFD), and not the physics, will be the basis for the proxy model.
![Page 7: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/7.jpg)
Paradigm Shift in Proxy Modeling
• Smart Proxy uses the pattern recognition capabilities of Artificial Intelligence and Data Mining (Big Data Analytics) to LEARN and then MIMIC the behavior of the original CFD.
• Smart Proxy is trained using massive amount of data that is generated from (only a handful of) CFD models runs and validated using blind CFD runs.
![Page 8: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/8.jpg)
Paradigm Shift in Proxy Modeling
• A single run of the original model is treated as a treasure trove of data and information that can be used to train and calibrate a new, smart model.
• This technology is the manifestation of “BIG DATA ANALYTICS” in engineering problem solving, specifically, in the numerical modeling arena.
![Page 9: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/9.jpg)
Surrogate Reservoir Modeling - SRMSMART PROXY MODEL FOR HIGH SPEED REPLICATION OF THE RESULTS OF NUMERICAL RESERVOIRSIMULATOR
![Page 10: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/10.jpg)
Numerical Model CharacteristicsStudy Completed 2005Implemented 2006Field Results Published(SPE 170664)
2014
Size of the Full Field Model 1MM Grid Blocks
Formation Type NaturallyFractured Carbonate
Number of Wells 167 (Horizontal)
Simulation Run 10 Hours on 12 parallel CPU
RESERVOIR SIMULATOR: ECLIPSESCHLUMBERGER
![Page 11: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/11.jpg)
11
Field Production Optimization StudyRESERVOIR SIMULATOR: ECLIPSESCHLUMBERGER
![Page 12: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/12.jpg)
12
RESERVOIR SIMULATOR: ECLIPSESCHLUMBERGER
Field Production Optimization Study
![Page 13: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/13.jpg)
13
PowersSRM
Oil
Rate
(bbl
s/yr
)
Cum. O
il Production (MM
bbls)
RESERVOIR SIMULATOR: POWERSSAUDI ARAMCO
Black Oil SimulationGreen Field6.5 MM Grid Blocks
![Page 14: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/14.jpg)
Oil
Rate
(bbl
s/yr
)
Cum. O
il Production (MM
bbls)
PowersSRM
RESERVOIR SIMULATOR: POWERSSAUDI ARAMCO
Black Oil SimulationGreen Field6.5 MM Grid Blocks
![Page 15: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/15.jpg)
Permeability Maps
History Matched Perm
High perm Low perm
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUPOtway CO2 Seq. Project, Australia
![Page 16: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/16.jpg)
Otway CO2 Seq. Project, Australia
16
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
Pressure Distribution
![Page 17: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/17.jpg)
Otway CO2 Seq. Project, Australia
17
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
Water Saturation Distribution
![Page 18: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/18.jpg)
SACROC CO2 EOR Project, Texas
18
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
-4,000-3,000
-2,000-1,000
0
-5,0
00-4
,000
-3,0
00-2
,000
-1,0
000
1,00
0
0.00 0.50 1.00 miles
0.00 0.50 1.00 km
File: input.irfUser: PNGE-UserDate: 4/2/2013Scale: 1:51166Y/X: 1.00:1Axis Units: m
11,538
14,622
17,707
20,792
23,877
26,961
30,046
33,131
36,215
39,300
42,385
Pressure (kPa) 2172-01-01 K layer: 18
The Coarse Reservoir Model
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000-4,000
-3,000-2,000
-1,0000
-5,0
00-4
,000
-3,0
00-2
,000
-1,0
000
1,00
0
0.00 0.50 1.00 miles
0.00 0.50 1.00 km
File: input.irfUser: PNGE-UserDate: 4/2/2013Scale: 1:51166Y/X: 1.00:1Axis Units: m
0.00
0.10
0.20
0.30
0.40
0.49
0.59
0.69
0.79
0.89
0.99
Gas Mole Fraction(CO2) 2172-01-01 K layer: 18
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
-4,000-3,000
-2,000-1,000
0
-5,0
00-4
,000
-3,0
00-2
,000
-1,0
000
1,00
0
0.00 0.50 1.00 miles
0.00 0.50 1.00 km
File: input.irfUser: PNGE-UserDate: 4/2/2013Scale: 1:51166Y/X: 1.00:1Axis Units: m
0.10
0.14
0.19
0.23
0.28
0.32
0.37
0.41
0.46
0.50
0.55
Water Saturation 2172-01-01 K layer: 18
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
-4,000-3,000
-2,000-1,000
0
-5,0
00-4
,000
-3,0
00-2
,000
-1,0
000
1,00
0
0.00 0.50 1.00 miles
0.00 0.50 1.00 km
File: input.irfUser: PNGE-UserDate: 4/2/2013Scale: 1:51159Y/X: 1.00:1Axis Units: m
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
Gas Saturation 2172-01-01 K layer: 18
![Page 19: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/19.jpg)
SACROC CO2 EOR Project, Texas
19
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
Water Saturation Distribution
Simulator (CMG) Results SRM Results % Relative Error# # # # # # # # # # # # # # 0 0 0 0# # # # # # # # # # # # # # # 0 0 0
# # ## # # # # # # # # # # # # # # # # # # # 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # 0.5 0.5 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.45 0.45 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.4 0.4 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.35 0.35 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.3 0.3 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.25 0.25 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.2 0.2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.15 0.15 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.1 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.05 0.05 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 0 0 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # ## # # # # # # # # # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Actual Data: Realization # 1, Layer= 18, 100 years after Injection, Feature= SW % SRM Data: Realization # 1, Layer= 18, 100 years after Injection, Feature= SW %
# # # # # 0 0 0 0# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 0 0 0
0.1 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 00.09 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.08 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.07 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.06 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.05 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Realization # 1, Layer= 18, 100 years after Injection, Feature= Absolute Error of SW %
# # # # # 0 0 0 0# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 0 0 0
0.1 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 00.09 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.08 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.07 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.06 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.05 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Realization # 12, Layer= 18, 100 years after Injection, Feature= Absolute Error of SW %
# # # # # 0 0 0 0# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 0 0 0
0.1 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 00.09 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.08 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.07 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.06 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.05 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Realization # 11, Layer= 18, 100 years after Injection, Feature= Absolute Error of SW %
# # # # # # # # # # # # # # 0 0 0 0# # # # # # # # # # # # # # # 0 0 0
# # ## # # # # # # # # # # # # # # # # # # # 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # 0.5 0.5 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.45 0.45 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.4 0.4 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.35 0.35 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.3 0.3 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.25 0.25 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.2 0.2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.15 0.15 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.1 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.05 0.05 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 0 0 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # ## # # # # # # # # # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Actual Data: Realization # 11, Layer= 18, 100 years after Injection, Feature= SW % SRM Data: Realization # 11, Layer= 18, 100 years after Injection, Feature= SW %# # # # # # # # # # # # # # 0 0 0 0# # # # # # # # # # # # # # # 0 0 0
# # ## # # # # # # # # # # # # # # # # # # # 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # 0.5 0.5 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.45 0.45 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.4 0.4 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.35 0.35 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.3 0.3 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.25 0.25 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.2 0.2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.15 0.15 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.1 0.1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.05 0.05 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 0 0 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # ## # # # # # # # # # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Actual Data: Realization # 12, Layer= 18, 100 years after Injection, Feature= SW % SRM Data: Realization # 12, Layer= 18, 100 years after Injection, Feature= SW %
![Page 20: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/20.jpg)
SACROC CO2 EOR Project, Texas
20
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
CO2 Mole Fraction Distribution
# # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # #
# # ## # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # 5000 5000 # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 4900 4900 # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4800 4800 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4700 4700 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4600 4600 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4500 4500 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4400 4400 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4300 4300 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4200 4200 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4100 4100 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 4000 4000 # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
# # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
SRM Data: Realization # 8, Layer= 18, 100 yrs after Injection, Feature= Pressure (psi)
Actual Data: Realization # 8, Layer= 18, 100 yrs after Injection, Feature= Pressure (psi)
# # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # #
# # ## # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # 5000 5000 # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 4900 4900 # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4800 4800 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4700 4700 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4600 4600 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4500 4500 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4400 4400 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4300 4300 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4200 4200 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4100 4100 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 4000 4000 # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
# # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
SRM Data: Realization # 9, Layer= 18, 100 yrs after Injection, Feature= Pressure (psi)
Actual Data: Realization # 9, Layer= 18, 100 yrs after Injection, Feature= Pressure (psi)
# # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # #
# # ## # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # 5000 5000 # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 4900 4900 # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4800 4800 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4700 4700 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4600 4600 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4500 4500 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4400 4400 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4300 4300 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4200 4200 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 4100 4100 # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 4000 4000 # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
# # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #
SRM Data: Realization # 10, Layer= 18, 100 yrs after Injection, Feature= Pressure (psi)
Actual Data: Realization # 10, Layer= 18, 100 yrs after Injection, Feature= Pressure (psi)
# # # # # 0 0 0 0# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 1 0 0
10 # # # # # # # # # # 0 0 0 0 1 0 0 0 0 0 0 0 0 09 # # # # # # # # # # 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 08 # # # # # # # # # # 0 0 0 1 0 0 0 0 1 1 0 1 0 0 1 0 1 0 0 0 0 07 # # # # # # # # # # 0 0 0 1 0 0 0 0 0 1 0 0 0 1 1 1 2 0 0 0 06 # # # # # # # # # # 1 0 0 0 1 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 05 0 # # # # # # # # # 0 0 0 0 1 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0 0 04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 03 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 01 # # # # # 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
# # # 0 1 2 0 0 0 0 0 0 0 0 0 0 0
Realization # 8, Layer= 18, 100 yrs after Injection, Feature= Relative Error %
# # # # # 0 1 1 0# # # # # # 1 1 1
# # # # # # # # # 0 0 1 0# # # # # # # # # # 1 1 1 0 2 0 0
10 # # # # # # # # # # 1 1 1 1 0 0 0 0 0 0 0 0 0 09 # # # # # # # # # # 1 1 1 0 1 1 0 0 0 0 0 0 0 1 1 0 0 0 08 # # # # # # # # # # 0 0 0 1 1 0 0 1 0 0 0 1 1 0 1 1 1 0 1 0 0 07 # # # # # # # # # # 0 1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 3 0 0 0 06 # # # # # # # # # # 1 0 0 0 0 0 1 1 0 0 0 1 1 1 1 1 1 0 0 0 0 05 1 # # # # # # # # # 1 0 0 0 2 1 0 0 0 0 1 1 0 0 1 0 0 0 0 0 0 04 1 1 1 1 1 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 03 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 02 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0
# # # 0 0 0 1 0 0 0 0 0 0 0 0 0 0
Realization # 9, Layer= 18, 100 yrs after Injection, Feature= Relative Error %
# # # # # 0 0 0 1# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 3 0 0
10 # # # # # # # # # # 0 0 0 0 1 0 0 0 0 0 0 0 0 19 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 08 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 07 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 1 0 0 06 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 05 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Realization # 10, Layer= 18, 100 yrs after Injection, Feature= Relative Error %
Simulator (CMG) Results SRM Results % Relative Error
![Page 21: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/21.jpg)
SACROC CO2 EOR Project, Texas
21
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
Pressure Distribution
Simulator (CMG) Results SRM Results Absolute Error# # # # # 0 0 0 0# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 0 0 0
0.1 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 00.09 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.08 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.07 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.06 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.05 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Realization # 9, Layer= 18, 9 years after Injection, Feature= Absolute Error %
# # # # # 0 0 0 0# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 0 0 0
0.1 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 00.09 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.08 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.07 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.06 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.05 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Realization # 10, Layer= 18, 100 years after Injection, Feature= Absolute Error %
# # # # # 0 0 0 0# # # # # # 0 0 0
# # # # # # # # # 0 0 0 0# # # # # # # # # # 0 0 0 0 0 0 0
0.1 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 00.09 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.08 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.07 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.06 # # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.05 0 # # # # # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.04 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.03 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.02 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00.01 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00 # # # # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
# # # 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Realization # 7, Layer= 18, 100 years after Injection, Feature= Absolute Error %
# # # # # # # # # # # # # # 0 0 0 0# # # # # # # # # # # # # # # 0 0 0
# # ## # # # # # # # # # # # # # # # # # # # 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # 0 0 0 0 0 1 1# # # ## # # # # # # # # # # # # # # # # # # # # # 1 1 # # # # # # # # # # 1 1 1 0 0 1 1 1 1 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.9 0.9 # # # # # # # # # # 0 0 0 1 1 0 0 1 1 0 0 0 1 1 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.8 0.8 # # # # # # # # # # 0 0 0 0 0 0 0 1 1 1 0 0 0 0 1 0 0 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.7 0.7 # # # # # # # # # # 0 0 0 0 0 0 1 1 1 1 0 1 1 1 1 1 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.6 0.6 # # # # # # # # # # 0 0 0 0 1 0 0 1 1 0 0 1 1 1 1 1 1 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.5 0.5 0 # # # # # # # # # 0 0 0 1 1 1 0 0 0 1 1 1 1 0 1 1 1 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.4 0.4 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 1 0 1 1 1 0 0 0 0 0 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.3 0.3 0 0 0 0 1 1 0 0 0 0 0 1 0 0 1 0 1 1 1 1 1 1 1 0 0 0 0 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.2 0.2 0 0 0 0 1 1 0 0 0 0 0 1 0 0 0 0 0 1 1 1 1 1 1 0 1 0 0 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.1 0.1 # # # # # 0 0 0 1 1 1 0 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 0 0 # # # # # 0 0 0 1 1 1 1 1 1 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0
# # ## # # # # # # # # # # # # # # # # # # 0 1 1 1 1 1 1 1 0 0 0 1 0 0
SRM Data: Realization # 10, Layer= 18, 9 years after Injection, Feature= CO2 Mole Fraction %
Actual Data: Realization # 10, Layer= 18, 9 years after Injection, Feature= CO2 Mole Fraction %
# # # # # # # # # # # # # # 0 0 0 0# # # # # # # # # # # # # # # 0 0 0
# # ## # # # # # # # # # # # # # # # # # # # 0 0 0 1# # # ## # # # # # # # # # # # # # # # # # # # # # # # # 0 0 0 0 0 1 1# # # ## # # # # # # # # # # # # # # # # # # # # # 1 1 # # # # # # # # # # 1 1 1 0 0 0 1 1 1 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.9 0.9 # # # # # # # # # # 0 0 0 1 1 1 0 1 1 1 1 1 1 1 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.8 0.8 # # # # # # # # # # 0 0 0 0 0 0 0 1 1 1 0 0 0 1 1 0 0 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.7 0.7 # # # # # # # # # # 0 0 0 0 0 0 1 1 1 1 0 1 1 1 1 1 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.6 0.6 # # # # # # # # # # 0 0 0 0 1 0 0 1 1 0 0 1 1 1 1 1 1 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.5 0.5 0 # # # # # # # # # 0 0 0 1 1 1 0 0 0 1 1 1 1 0 1 1 1 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.4 0.4 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 1 0 1 1 1 0 0 0 0 0 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.3 0.3 0 0 0 0 1 1 0 0 0 0 0 1 0 0 1 0 1 1 1 1 1 1 1 0 0 0 0 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.2 0.2 0 0 0 0 1 1 0 0 0 0 0 1 0 0 0 0 0 1 1 1 1 1 1 0 1 0 0 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.1 0.1 # # # # # 0 0 0 1 1 1 0 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 0 0 # # # # # 0 0 0 1 1 1 1 1 1 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0
# # ## # # # # # # # # # # # # # # # # # # 0 1 1 1 1 1 1 1 0 0 0 1 0 0
SRM Data: Realization # 9, Layer= 18, 9 years after Injection, Feature= CO2 Mole Fraction %
Actual Data: Realization # 9, Layer= 18, 9 years after Injection, Feature= CO2 Mole Fraction %
# # # # # # # # # # # # # # 0 0 0 0# # # # # # # # # # # # # # # 0 0 0
# # ## # # # # # # # # # # # # # # # # # # # 0 0 0 1# # # ## # # # # # # # # # # # # # # # # # # # # # # # # 0 0 0 0 0 1 1# # # ## # # # # # # # # # # # # # # # # # # # # # 1 1 # # # # # # # # # # 1 1 1 0 0 0 1 1 1 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.9 0.9 # # # # # # # # # # 0 0 0 0 1 0 0 1 1 1 1 1 1 1 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.8 0.8 # # # # # # # # # # 0 0 0 0 0 0 0 1 1 1 0 0 1 0 1 1 1 1 1 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.7 0.7 # # # # # # # # # # 0 0 0 0 0 0 1 1 1 1 0 1 1 1 1 1 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.6 0.6 # # # # # # # # # # 0 0 0 0 1 0 0 1 1 0 0 1 1 1 1 1 1 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.5 0.5 0 # # # # # # # # # 0 0 0 1 1 1 0 0 0 1 1 1 1 0 1 1 1 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.4 0.4 0 0 0 0 0 0 0 0 1 0 0 0 0 0 1 0 0 1 0 1 1 1 0 0 0 0 0 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.3 0.3 0 0 0 0 1 1 0 0 0 0 0 1 0 0 0 0 1 1 1 1 1 1 1 0 0 0 0 1 1 1 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.2 0.2 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 0 1 0 0 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # 0.1 0.1 # # # # # 0 0 0 1 1 1 0 1 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0# # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # 0 0 # # # # # 0 0 0 1 1 1 1 1 1 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0
# # ## # # # # # # # # # # # # # # # # # # 0 1 1 1 1 1 1 1 0 0 0 1 0 0
SRM Data: Realization # 7, Layer= 18, 9 years after Injection, Feature= CO2 Mole Fraction %
Actual Data: Realization # 7, Layer= 18, 9 years after Injection, Feature= CO2 Mole Fraction %
![Page 22: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/22.jpg)
SACROC CO2 EOR Project, Texas
22
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
High Resolution Model
Permeability
Porosity
![Page 23: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/23.jpg)
SACROC CO2 EOR Project, Texas
23
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
High Resolution Model
![Page 24: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/24.jpg)
SACROC CO2 EOR Project, Texas
24
RESERVOIR SIMULATOR: AMEXCOMPUTER MODELING GROUP
High Resolution Model
![Page 25: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/25.jpg)
Smart Proxy for CFDCAN SMART PROXY BE SUCCESSFULLY APPLIED TO CFD?
![Page 26: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/26.jpg)
A Multi-phase CFD Model
• Initial pressure is 1 atm. in the entire domain1 atm. = 101,325 Pa
Grid Classification I, j, k - Cell size No. of Cell No. of Grid
Coarse8 x 48 x 8(15 mm)
3,072 3,969
Medium12 x 72 x 12
(10 mm)10,368 12,337
Fine18 x 108 x 18
(6.6 mm)34,992 39,349
Very Fine27 x 162 x 27
(4.4 mm)118,098 127,792
![Page 27: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/27.jpg)
Three different physics
Early-timeTime step 200
Mid-timeTime step 1000
Late-timeTime step 4000
![Page 28: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/28.jpg)
Early-timeTime step 200
Five cross-sections used to show details in 2-D
K = 1
K = 14
K = 7
K = 27
K = 21
![Page 29: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/29.jpg)
Mid-timeTime step 1000
K = 1
K = 14
K = 7
K = 27
K = 21
![Page 30: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/30.jpg)
Late-timeTime step 4000
K = 1
K = 14
K = 7
K = 27
K = 21
![Page 31: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/31.jpg)
Data from the CFD Runs
Data from the CFD Runs
Spatio-Temporal Database
Generated to be used by the
Smart CFD Proxy
Data Smart CFD Proxyused to Build
Key factors of multi-phase gas-solid systemVolume fraction (ɛ) Gas Density (ρ)
Velocity vector of gas (u, v, w) Particle diameter (d)
Velocity vector of solid (u, v, w) Maximum packing volume fraction (ɛ*)
Pressure field of gas (P) Location to the boundaries
Pressure field of solid(P) Location to the interface(x, y, z)
Time (t) Viscosity (μ and ʎ)
Gravity force (g)
![Page 32: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/32.jpg)
Displaying Smart CFD Proxy
• Results are displayed in 5 different cross-sections:
• K = 1, 7, 14, 21, 27
K = 1
K = 14
K = 7
K = 27
K = 21
X = 27Y = 162Z = 27
![Page 33: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/33.jpg)
![Page 34: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/34.jpg)
![Page 35: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/35.jpg)
Thank you
![Page 36: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/36.jpg)
Extra Slides
![Page 37: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/37.jpg)
Part One – Non-Cascading Results
• In Non-Cascading scheme at every time-step actual CFD results are used as input and the trained Smart Proxy forecasts the results for the next time-step.
• No error accumulation takes place in this scheme.
• This scheme, as the first step, explores the learning capabilities of the Smart Proxy.
nC
![Page 38: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/38.jpg)
Early Time Results
• Smart Proxy is generated for the early time in the system when the bubble behavior is reasonably calm and non-chaotic.
nC
![Page 39: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/39.jpg)
Results of Layer 1 – Gas Fraction
• Model was trained using Time-step 101 as input and Time-step 102 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 102 through Time-step 126
• Results generated by the Smart Proxy are compared with actual CFD output.
nC
![Page 40: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/40.jpg)
nC
![Page 41: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/41.jpg)
Results of Layer 27 – Gas Fraction
• Model was trained using Time-step 101 as input and Time-step 102 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 102 through Time-step 126
• Results generated by the Smart Proxy are compared with actual CFD output.
nC
![Page 42: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/42.jpg)
nC
![Page 43: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/43.jpg)
Multiple bubbles in the system
• The process is now repeated when several bubbles have already been formed in the system.
nC
![Page 44: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/44.jpg)
Results of Layer 1 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4001 through Time-step 4040
• Results generated by the Smart Proxy are compared with actual CFD output.
nC
![Page 45: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/45.jpg)
nC
![Page 46: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/46.jpg)
Results of Layer 7 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4001 through Time-step 4040
• Results generated by the Smart Proxy are compared with actual CFD output.
nC
![Page 47: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/47.jpg)
nC
![Page 48: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/48.jpg)
Results of Layer 14 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4001 through Time-step 4040
• Results generated by the Smart Proxy are compared with actual CFD output.
nC
![Page 49: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/49.jpg)
nC
![Page 50: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/50.jpg)
Results of Layer 21 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4001 through Time-step 4040
• Results generated by the Smart Proxy are compared with actual CFD output.
nC
![Page 51: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/51.jpg)
nC
![Page 52: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/52.jpg)
Results of Layer 27 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4001 through Time-step 4040
• Results generated by the Smart Proxy are compared with actual CFD output.
nC
![Page 53: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/53.jpg)
nC
![Page 54: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/54.jpg)
Part Two – Cascading Results
• In Cascading scheme at every time-step results from the Smart Proxy are used as input and the trained Smart Proxy forecasts the results for the next time-step. (Smart Proxy receives feedback from itself)
• Models are developed for all output variables in CFD such as Void Fraction, and gas/solid Pressure and Velocity.
• In this scheme the error has the opportunity to accumulate.
• Limits of the information content of each time-step is explored in this scheme.
C
![Page 55: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/55.jpg)
Early Time Results
• Smart Proxy is generated for the early time in the system when the bubble behavior is reasonably calm and non-chaotic.
C
![Page 56: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/56.jpg)
Results of Layer 1 – Gas Fraction
• Model was trained using Time-step 101 as input and Time-step 102 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 101 through Time-step 105
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 57: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/57.jpg)
C
![Page 58: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/58.jpg)
Results of Layer 7 – Gas Fraction
• Model was trained using Time-step 101 as input and Time-step 102 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 101 through Time-step 105
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 59: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/59.jpg)
C
![Page 60: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/60.jpg)
Results of Layer 14 – Gas Fraction
• Model was trained using Time-step 101 as input and Time-step 102 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 101 through Time-step 105
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 61: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/61.jpg)
C
![Page 62: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/62.jpg)
Results of Layer 21 – Gas Fraction
• Model was trained using Time-step 101 as input and Time-step 102 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 102 through Time-step 126
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 63: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/63.jpg)
C
![Page 64: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/64.jpg)
Results of Layer 27 – Gas Fraction
• Model was trained using Time-step 101 as input and Time-step 102 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 101 through Time-step 105
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 65: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/65.jpg)
C
![Page 66: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/66.jpg)
Multiple bubbles in the system
• The process is now repeated when several bubbles have already been formed in the system.
C
![Page 67: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/67.jpg)
Results of Layer 1 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4002 through Time-step 4014
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 68: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/68.jpg)
C
![Page 69: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/69.jpg)
Results of Layer 7 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4002 through Time-step 4010
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 70: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/70.jpg)
C
![Page 71: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/71.jpg)
Results of Layer 14 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4002 through Time-step 4024
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 72: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/72.jpg)
C
![Page 73: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/73.jpg)
Results of Layer 21 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4002 through Time-step 4020
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 74: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/74.jpg)
C
![Page 75: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/75.jpg)
Results of Layer 27 – Gas Fraction
• Model was trained using Time-step 4001 as input and Time-step 4002 as output.
• Trained on 75% of data and validated using the remaining 25%
• Trained model was deployed in forecast mode and generated Time-step 4002 through Time-step 4012
• Results generated by the Smart Proxy are compared with actual CFD output.
C
![Page 76: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/76.jpg)
C
![Page 77: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/77.jpg)
Paradigm Shift in Proxy Modeling
• VERY LARGE AMOUNTS OF DATA is carefully extracted from a handful of runs of the original model.
• The process of data extraction from several runs of the original model is ENGINEERED such that it can be used to train, calibrate and validate a smart, new model.
![Page 78: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/78.jpg)
Paradigm Shift in Proxy Modeling
• The Smart Proxy model is developed using MACHINE LEARNING that is now the main paradigm for treating and handling “BIG DATA”.
• The Smart Proxy learns to mimic the behavior of the original model with all its complexities and intricacies.
• Once developed, the Smart Proxy Model runs at very low computational cost (thousands of simulation runs in minutes).
![Page 79: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/79.jpg)
The Spatio-Temporal Database
D1 D2
D4
D3
D5
D6
Tier One Neighbor Cells
Number of inputs 6 + 9 + 6 x 9 = 69Number of outputs = 9
![Page 80: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/80.jpg)
![Page 81: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/81.jpg)
Optimize the number of inputs
![Page 82: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/82.jpg)
Tornado chart by adding all the weights
0
5
10
15
20
25
30
35
Tornado Chart of Neural Network Weights
![Page 83: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/83.jpg)
Eliminating 14 parameters
![Page 84: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/84.jpg)
RSME
![Page 85: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/85.jpg)
![Page 86: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/86.jpg)
The Tier System
![Page 87: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/87.jpg)
The Tier System
Tier OneInterface: Plane
Tier TwoInterface: Line
Tier ThreeInterface: Point
![Page 88: REPLICATING ESULTSACCURATELY ATHIGHSPEED · 2020. 8. 13. · Outline • Smart Proxy • Engineering Application of Big Data Analytics • Original Application in Reservoir Modeling](https://reader036.fdocuments.us/reader036/viewer/2022081621/611db35855997d308640d9e9/html5/thumbnails/88.jpg)
Tier One - Subsystem
Tier OneInterface: Plane
Tier OneTwo Subsystems
Tier 1Input and Tier 1Output
Tier OneThree Subsystems
Tier 1EW , Tier 1SN , and Tier 1BA
Tier 1EW
Tier 1SNTier 1BA
Tier 1Input
Tier 1Output