Professional Documents
Culture Documents
Alvarado 2002
Alvarado 2002
of the risks involved. This is a consequence of uncertainties, Carvajal et al.3 describe the management approach for
due to incomplete information, a constant in the E & P technology evaluation in the IFL's as follows:
business. Methods for analysis should be designed to enable • Advanced reservoir characterization, simulation
iteration during evaluations, including progressively more and visualization.
details, such that they allow us to refine as the opportunity • Drainage strategies, combining EOR methods and
becomes more attractive and data gathering (reservoir well architectures.
characterization mostly) turns out justifiable. Along with these • Speedy well construction, with minimum damage
ideas, PDVSA has developed the concept of the Integrated and cost.
Field Laboratory (IFL) to facilitate testing field technologies • IOR technologies to deal with injectivity and
and their deployment in a number of exploitation units. productivity enhancement.
A data mining strategy applied to a collated database of • Advanced monitoring technologies.
international project results is used here for knowledge • Production fluid handling technologies.
extraction on applicability of EOR processes. Statistical • Risk evaluation and mitigation.
analysis of the data yields importance of variables, in terms of
Eight pilot tests have been planned since 1996. As seen in
how they influence clustering of reservoirs. A small number of
this section, intensive application of technologies in field
these variables, representing average values for each reservoir operations is an important part of the objective in the IFL
are used to rank EOR processes and extract rules. It is
strategy. However, technology evaluation of that level of
important to notice that, as mentioned previously, the process
detail is only possible in pilot areas. Screening criteria that
of inquiring the database does not end at the first include the possibility of portafolio analysis is an answer to
representation of the data, which means that further refinement
extrapolation of lessons learned in pilot tests. Alternative
is necessary, until a decision can be made or information is
approaches, based on artificial intelligence, are now describe
exhausted from the extraction process. This differs from
to introduce the context of this work.
traditional analysis in the sense that several iterations of
the screening/ranking process are not only possible, but
Artificial Intelligence
also necessary.
Artificial Intelligence (AI), specifically Neural Networks,
The paper is organized as follows. After this introduction,
Fuzzy Logic, and Expert System, have been often proposed
a summary of the IFL strategy is summarized. Then, reference and used for supporting E&P operations. Their use varies
to artificial intelligence methods and several EOR screening
depending on the specific problem. In the case of neural
methodologies is carried out. The proposed methodology is
network and fuzzy logic, they have been proposed for data
then explained, followed by the results section. Closing filtering11 (smoothing), or as modeling tools. All the potential
remarks and recommendations are provided at the end. This
of these information processing systems are used to build non-
work does not pretend to be comprenhensive, but rather
linear models for oil production forecast, log interpretation to
intends to show a first view of a whole strategy thought of for identify total porosity as well as lithofacies12, or reservoir
these purposes.
property related estimations13. In the case of Expert system is
mostly used for knowledge representation in the form of IF-
Integrated Field Laboratory THEN rules, where specific Know-How from Experts are used
The idea behind the IFL philosophy is the speedy
to build schemes that would be automated and used for
evaluation and incorporation of technologies to field
modeling an Expert reasoning. Some expert system have been
operations3. However, finding the best technology for developed recently14,15, for different disciplines of E&P, which
individual reservoirs would represent an endless task. Here
included drilling areas16,17, well bore simulation18,19, well
comes in the idea of grouping reservoirs by type, i.e. by
testing and logging20,21, EOR and fluid property
analyzing together reservoirs with similar characteristics.
predictions22,23. It is important to mention that combinations of
Applicability criteria for EOR technologies and reservoir types
these different tools are also possible, i.e. fuzzy rules are used
are a motivation for investing in these advanced pilot test
to increase the capability of the expert system to deal with
areas. One of the weakest aspects in IFL projects, has been the
uncertainties. Also, fuzzy activation functions are used in
method for extrapolation of learned strategies from a pilot area
combination with Neural Network topologies so that some
to a large set of reservoirs. Several EOR technologies are form of regression techniques to adjust the fuzzy set rules are
under scrutiny in the IFL projects: Optimized Water flooding3
possible. A recent new player in these AI solution suites is
(OWF), Alkaline-Surfactant-Polymer (ASP) flooding4-6,
Machine Learning (ML)24. One possible realization of this
Continuous Steam Floding (CSF) in heavy oil7 and Water- technique is the combination of Clustering techniques25,26 and
Alternating-Gas (WAG) injection8-10. Fig. 1 illustrates the
rule extraction algorithms27. In this approach, all the data
typical workflow for evaluation in an IFL, for the
available is used to extract implicit and explicit process or
VLE example. business rules hidden within the data, whether heuristic or
not28. In all these cases, AI has showed an excellent
performance as well as simplicity in the final solution. With
these techniques, the probabilities of success are strongly
SPE 78332 SELECTION OF EOR/IOR OPPORTUNITIES BASED ON MACHINE LEARNING 3
In addition to the methods statistics that can be applied to as the clustering algorithms are flexible enough to complete
each cluster, we have developed a set of rules that allow us to this task. Rules derived from the automatic extraction method
characterize each one of the clusters. The set of rules should , from our point of view, be complemented by using
associated to clusters shown in the projection of Fig.3 are expert opinions, guided in turn by the reduced representation.
listed in Table 4. It is important to highlight that not all six
variables are in principle used to define a rule automatically. Conclusions
This is a consequence of the algorithm employed for this 1. Space reduction techniques have been applied
purpose. However, it has been already forseen that rules can successufully to produce bidimensional maps that clearly
be, and will be, complemented by expert opinion, based on show reservoir types by using 6 reservoir variables.
observation of the maps. 2. The generated maps allowed us to establish applicability
To find reservoir typologies, representative candidates for criteria or selection rules, based on international
specific processes, that is, clusters made out of different experience on EOR processes.
reservoirs to which the same EOR method has been applied, 3. Several Venezuelan reservoirs were mapped, based on
each of the six clusters in Fig. 3 are reanalyzed (the same their reported average reservoir variables, and sensible
applies to other clusters). This is part of the possible conclusions on applicability of EOR processes were
refinement of the data analysis, mentioned before. The latter drawn from the analysis method proposed here.
means that the same methodology and cluster algorithms can 4. The outcomes of this work drives further development of
be used over the subset of data integrated for each one of the the techniques proposed here to refine the
reservoirs that conform cluster 5 or any other cluster (see Fig. screening/ranking criteria based on detailed analysis on
3). For example, looking closely at cluster 5 (Fig. 4), we see the available data.
once again a new classification (six clusters A, B, C, D, E and 5. Firmer rules and conclusions can be drawn as the
F). We also observe that pure typologies, those related to one gathered experience, represented in the database, is
method only, are better defined. Cluster B and cluster C are enlarged. This, however, would require collaboration
clear examples of pure typologies of thermal methods within from oil companies, as the results of the application of
global cluster 5. Since the new analysis is performed on a EOR methods are not often found in the open litterature.
reduced set of data of the international database, new results
and a different set of rules are obtained. Acknowledgements
Figure 5 shows a projection of the international database We would like to thank PDVSA-Intevep for permission to
that includes two Venezuelan reservoirs. For instance, publish this paper.
Reservoir A (a relatively shallow light oil reservoir in southern
Venezuela- indicated by the dashed Brown Circle) is located References
in cluster 4. On the other hand, reservoir B (a deep extra- 1. Goodyear, S. G. and Gregory, A. T.: “Risk Assessment and
heavy oil reservoir in western Venezuela – marked with the Management in IOR Projects” Paper SPE 28844 presented at the
dashed Red Circle) is located in cluster 5. For reservoir B, the 1994 European Petroleum Conference, London, October 25-27.
set of points located within the dotted red circle corresponds to 2. Thompson, M. A. and Goodyear, S. G.: “Identifying Improved Oil
Recovery Potential: A New Systematic Risk Management”
the sensitivity values for some of the process variables in the Paper SPE 72103 presented at the 2001 Asia Pacific Improved
same reservoir. This leads to the possibility that sensitivity Oil Recovery Conference, Kuala Lumpur, October 8–9.
analysis on reservoir history can be explored with the 3. de Carvajal, G., Velasquez, A., Graterol, J., Ramirez, F., and
proposed methodology Medina, M.: “Lagomar's Integrated Field Laboratory for
Reanalyzing cluster 4, which contains the Reservoir A (Fig. Intensive Evaluation of Technologies” Paper SPE 53984
5), it can be concluded that the statistics and list of methods that presented at the 1999 SPE Latin American and Caribbean
can be applied to this reservoir are those listed in Table 6. Petroleum Engineering Conference, Caracas, April 21–23.
If we take a look at reservoir B (Fig. 6 - Dotted Red 4. Manrique, E., de Carvajal, G., Anselmi, L., Romero, C., Chacón,
Circle), we can see that the reservoir belongs to cluster 5 in the L.: "Alkali/Surfactant/Polymer at VLA 6/9/21 Field in
Maracaibo Lake: Experimental Results and Pilot Project
general projection map. This means that within a global scope, Design" Paper SPE 59363 presented at the 2000 SPE/DOE
the reservoir has conditions for thermal methods. However, Improved Oil Recovery Symposium held in Tulsa, April 3–5.
notice that the original reservoir and its sensitivity values are 5. Hernández, C., Chacón, L J., Anselmi, L,, Baldonedo, A,, Qi, J., and
located on the boundaries of the cluster. However, a zoom of Pitts, M. J.: " ASP System Design for an Offshore Application in
cluster 5 (Fig. 7) shows that reservoir B and its sensitivity the La Salina Field, Lake Maracaibo" Paper SPE 69544 presented
values are close to cluster F, but out of the region that defines at the 2001 Latin American and Caribbean Petroleum Engineering
the cluster. The latter means that by making operational Conference, Buenos Aires, March 25–28.
changes such as decreasing the reservoir’s pressure, reservoir 6. Hernández, C., Chacón, L., Anselmi, L., Angulo, R., Manrique,
B can move into cluster F. E., Romero, E., de Audemard, N., Carlisle, C.: “Single Well
Chemical Tracer Test to Determine ASP Injection Efficiency at
The statistics associated with this reservoir are as shown Lagomar VLA-6/9/21 Area, C4 Member, Lake Maracaibo,
by Table 6 and the set of rules defined for cluster 5 are those Venezuela” Paper SPE 75122 presented at the 2002 Improved
in Table 7. The whole procedure applied for the examples Oil Recovery Symposium, Tulsa, April 13–17.
discussed here can be used if more variables were considered,
SPE 78332 SELECTION OF EOR/IOR OPPORTUNITIES BASED ON MACHINE LEARNING 5
7. Lopez, E, Rojas, L., Mata, T., Mendoza, H., and Briñez, A.: " 22. Gharbi, R.B. 2000. An Expert System for Selecting and
Integrated Laboratory Field Application or Thermal Recovery Designing EOR Processes. Accepted for Publication, J.Petrol.
Process" Paper SPE 53983 presented at the 1999 SPE Latin Sci.Eng.March 2000.
American and Caribbean Petroleum Engineering Conference, 23. Dharan, M.B., Turek, E.A. and Vogel, J.L.: “The fluid properties
Caracas, April 21–23. measurement expert system” Paper SPE 19134 presented at the
8. Manrique, E., Calderón, G., Mayo, L., and Stirpe, M. T.: "Water- 1989 Petroleum Computer Conference, San Antonio, June 26-28.
Alternating-Gas Flooding in Venezuela: Selection of Candidates 24. Mitchell, T. M., “Machine Learning”, Mc Graw Hill, 1997.
Based on Screening Criteria of International Field Experiences" 25. Ranson, A.; Hernandez, K.; Matheus, J.; Vivas A.: “Monitoring
Paper SPE 50645 presented at the 1998 European Petroleum and knowledge extraction in real time multivariable dynamical
Conference, The Hague, October 20-22. processes”. ANNIE 2001. 2001.
9. Manrique, E. , Padrón, R., Surguchev, L., De Mena, J., McKenna, 26. Rujano, R.: Implementation and evaluation of clustering
K.: "VLE WAG Injection Laboratory Field in Maracaibo Lake" algorithms using non-deterministic search algorithms to find
Paper SPE 65128 presented at the 2000 European Petroleum the optimal number of classes, Thesis PDVSA INTEVEP, 2002.
Conference, Paris, October 24-25. 27. Quinlan, J. R : C4.5 Programs for Machine Learning, Morgan
10. Alvarez, C., Manrique, E., Alvarado, V., Samán, A., Surguchev, Kaufman Publishers, California 1993.
L., and Eilertsen, T.: "WAG pilot at VLE field and IOR 28. Ranson, A., Hernández, K.Y., Matheus, and Vivas A.A.: “A New
opportunities for mature fields at Lake Maracaibo" Paper SPE Approach to Identifying Operational Conditions in
72099 presented at the 2001 Asia Pacific Improved Oil Multivariable Dynamic Processes Using Multidimensional
Recovery Conference, Kuala Lumpur, October 8–9. Projection Techniques” Paper SPE 69523 presented at the 2001
11. Balch, R. S., Hart, D. M., Weiss, W. W., and Broadhead R. F.: Latin American and Caribbean Petroleum Engineering
“Regional Data Analysis to Better Predict Drilling Success: Conference, Buenos Aires, March 25–28.
Brushy Canyon Formation, Delaware Basin, New Mexico” 29. Joseph, J., Taber, F., David, M. and Seright R. S.: “EOR
Paper SPE 75145 presented at the 2002 Improved Oil Recovery Screening Criteria Revisited” Paper SPE/DOE 35385 presented
Symposium, Tulsa, April 13-17. at the 1996 SPE/DOE Symposium on Improved Oil Recovery,
12. Weiss, W. W., Balch, R. S., and Stubbs B. A.: “How Artificial Tulsa, April 21-24.
Intelligence Methods Can Forecast Oil Production” Paper SPE 30. Thomas, S., and Farouq, A.: “Field Experience with Chemical Oil
75143 presented at the 2002 Improved Oil Recovery Recovery Methods”, Chemical Abstract, Vol. 3, 1995, P:45-3
Symposium, Tulsa, April 13-17. to 49-3..
13. Surguchev, L. and Li, L.: “IOR Evaluation and Applicability 31. Rao, D: Gas Injection EOR a New Meaning in the New
Screening Using Artificial Neural Networks” Paper SPE 59308 Millennium. The Journal of Canadian Petroleum Technology,
presented at the 2000 SPE/DOE Improved Oil Recovery Feb. 2001, Volume 40, No. 2.
Sympsium held in Tulsa, April 3-5. 32. Dobitz, J. K. and Prieditis J.: “A Stream Tube Model for the PC”
14. Bergen, J. K. and Hutter, J.E. 1986. The Mudman Service-an SPE/DOE 27750 presented at the 1994 Symposium on
artificial intelligence aid for drilling. DrillIng and Production Improved Oil Recovery, Tulsa, April 17-20.
Technoly Symposium PD-vol. 4 (Book No. 100203), The 33. K. Thukral and M. Karuppasamy. Hydrocarbon Development -
American Society of Mechanical Engineering. Simulation of EOR Applications. Energy, Vol. 16, No. 9, pp.
15. Peden, J.M. and Tovar, J.J.: “Sand prediction and exclusion 1207-1212. 1991.
decision support using an expert system” Paper SPE 23165 34. T. Okazawa, P. E. Bozac, A. C. Seto, G. R. Howe. Analytical
presented at the 1991 Offshore European Conference, Aberdeen, Software for Pool-wide Performance Prediction of EOR
September 3-6. Processes. The Jorunal of Canadian Petroleum Technology,
16. Einstein, E.E. and Edwars, K.W. 1990. Comparison of an expert April 1995, Vol. 34, No. 4.
system to human experts in well-log analysis and interpretation. 35. PRIze: Analytical Model for Evaluating the EOR Potential of
SPE Form. Eval. March: 39-45. Petroleum Reservoir. 1994.
17. Allain, O. and Houze, O.P.: “A Practical artificial intelligence 36. Yu, J. P., Zhuang, Z., and Watts, R. J: “Microcomputer
application in well testing interpretation” Paper SPE 24287 Applications in Economic Assessment and Risk Analysis of
presented at the 1992 European Petroleum Conference, CO2 Miscible Flooding Process” Paper SPE 19318 presented at
Stavanger, May 25-27. the 1989 Eastern Regional Meeting, West Virginia, Oct. 24-27.
18. Hutchin, L.A., Burton, R.K. and Macintosh, D.J.: “An expert 37. Basnieva, I. K., Zolotukhin, A. B., Eremin, N. A., and Udovina.
system for analyzing well performance” Paper SPE 35705 E. F.: “Comparative Analysis of Successful Application of EOR
presented at the 1996 Western Regional Meeting, Anchorage, in Russia and CIS” Paper SPE 28002 presented at the 1994
May 22-24. University of Tulsa Centennial Petroleum Eng. Symp, Tulsa
19. Alegre,L., Morokooka, C.K. and Rocha, A.F.: “Intelligence August 29-31.
Diagnosis of rod pumping problems” Paper SPE 26516 38. Chung, T.-H., Carroll, H. B, and Lindsey, R.: “Application of
presented at the 1993 Annual Technical Conference, Houston, Fuzzy Expert Sistems for EOR Project Risk Analysis” Paper
October 3-6. SPE 30741 presented at the 1995 Annual Technical Conference
20. Patricio, A.R., Rocha, A.F. da and Morooka, C.K.: “Seplant: an & Exhibition, Dallas, October 22-25.
expert system for process plant and gas lift well” Paper SPE 39. Gharbi, R. B. C.: “An expert system for selecting and designing
28238 presented at the 1994 Petroleum Computer Conference, EOR processes”. Journal of Petroleum and Engineering, 27
Dallas, July 31-August 3. (2000) 33-47.
21. Corpoven, M.V.O.: “Real Time Expert System (R.E.T.S) for rod 40. Abou-Kassem, J. H. Screening of Oil Reservoirs for selecting
pumping optimization” Paper SPE 34185 presented at the 1995 Candidates of Polymer Injection. Energy Sources, 21:
Petroleum Computer Conference, Houston, June 11-14. 5-16, 1999.
6 V. ALVARADO, A. RANSON, K. HERNÁNDEZ, E. MANRIQUE, J. MATHEUS, T. LISCANO, N. PROSPERI SPE 78332
29,35,40
Table 1. Summary of Screening Criteria for Polymer and CO2 Flooding
Parameter Polymer CO2
Flooding Flooding
Oil Gravity (°API) > 22 > 25
Oil viscosity(cp) < 100 < 15
Crude Oil Composition NC High % C5-C12 fraction
Oil saturation (% PV) > 50 > 25
Water salinity (ppm) < 100000 NC
Water hardness (ppm) < 5000 NC
Mobility ratio 2 - 40 NC
Reservoir temperature (°F) < 200 NC
Rock type Sandstone preferred Sandstone or carbonate
Permeability (mD) > 50 NC
Depth (ft) < 9000 > 2500
Net thickness NC Wide range
Minimum Miscibility Pressure NC < Original pressure
Drive mechanism No gas cap and no bottom No gas cap
water drive
Table 5. Statistics of recovery methods associated with Cluster 4-C (see Fig. 6).
Method %
Polymer 33.33
CO2 Misc. 16.67
Steam 16.67
Air 16.67
Water flooding 16.67
Table 7. Set of Rules defined for New Analysis Cluster 5 ( Reservoir B).
Cluster Number RULE
1 IF IF VISC <= 135 && API > 24.5 && PRESS <= 489.85 Then Cluster 1
4 IF API <= 20.05 && TEMP > 107.5 &&PERM > 314.75 &&VISC > 95
Then Cluster 4
5 IF VISC <= 135 && API > 20.05 &&API <= 24.5 Then Cluster 5
6 IF API <= 20.05 && TEMP > 107.5 &&PERM <= 314.75 5 Then Cluster
6
8 V. ALVARADO, A. RANSON, K. HERNÁNDEZ, E. MANRIQUE, J. MATHEUS, T. LISCANO, N. PROSPERI SPE 78332
Preliminary
Main
Economy Laboratory
Screening
Evaluation evaluation
Criteria
(EE)
SPE
65128
SPE 50645
Simulation studies
and Pilot Design
Unsuccessful
New drainage
strategies
Project SPE
Evaluation Project performance
72099
and and
75259
EE Monitoring
Large field scale
application Successful
Fig. 1. Typical planning for IFLs, illustrated through the VLE example. SPE papers reflect the progress of this particular case.
Wag HC Misc.
5%
N2 Inmisc.
4%
Wag CO2 Inmisc.
CO2 Inmisc. 1%
5% N2 Misc. Wag HC Inmisc. Water Flooding
Wag CO2 Misc. 5% 1%
Polymer
7% Wag N2 Misc. Air
Steam 1%
CO2 Misc.
7% Steam
Wag CO2 Misc.
CO2 Inmisc.
N2 Misc.
CO2 Misc. Water Flooding Wag HC Misc.
9% 30% N2 Inmisc.
Air Wag CO2 Inmisc.
10% Polymer Wag HC Inmisc.
15% Wag N2 Misc.
Fig. 2. Distribution of EOR Methods reported in the collated database. Waterflooding abunts in the database, followed by polymer flooding.
Scarce data were available for processes such as Nitrogen injection
SPE 78332 SELECTION OF EOR/IOR OPPORTUNITIES BASED ON MACHINE LEARNING 9
Cluster 4
Cluster 6
Method %
CO2 Immisc. 22.58 Method %
Air 12.90 N2 Misc. 42.86
Cluster 5 Water Flooding 12.90 N2 Inmisc. 21.43
CO2 Misc. 9.68
Method % WAG N2 Misc. 14.29
Polymer 9.68
Air 41.38 WAG HC Inmisc. 9.68 Cluster 6 Water Flooding 14.29
Steam 27.59 N2 Misc. 6.45 WAG HC Misc. 7.14
CO2 Immisc. 10.34 WAG HC Misc. 6.45
N2 Inmisc. 3.23
Polymer 8.62 Steam 3.23
WAG CO2 Inmisc. 5.17 WAG CO2 Misc. 3.23
Water Flooding 5.17
N2 Inmisc. 1.72 CO2 Inmisc.
CO2 Misc.
N2 Inmisc.
N2 Misc.
Polymer
Steam
Wag CO2 Inmisc.
Cluster 2
Wag HC Inmisc.
Method %
W ater Flooding 38.46 Wag CO2 Misc.
W AG CO2 Misc. 13.46 Wag HC Misc.
W AG HC Misc. 13.46
N2 Misc. 9.62 Air
CO2 Misc. 7.69 Water Flooding
N2 Inmisc. 7.69
Cluster 1 Polymer 5.77
Method % Air 3.85
W ater Flooding 29.17 Cluster 3
CO2 Misc. 20.83
Method %
Polymer 18.75
N2 Inmisc. 6.25
Water Flooding 48.28
Steam 6.25 Polymer 25.29
WAG HC Misc. 6.25 WAG CO2 Misc. 12.64
CO2 Immisc. 4.17 CO2 Misc. 10.34
WAG CO2 Misc. 4.17 N2 Inmisc. 1.15
N2 Misc. 2.08 WAG HC Misc. 1.15
WAG N2 Misc. 2.08 Steam 1.15
CO2 Inmisc.
N2 Inmisc.
Polymer
Cluster D Steam
Method % Air
CO2 Inmisc. 33.33 Water Flooding
Air 22.22 Wag CO2 Inmisc.
Wag CO2 Inmisc. 22.22
Water Flooding 11.11
Polymer 11.11
Cluster F
Method %
CO2 Inmisc. 40
N2 Inmisc. 20
Wag CO2 Inmisc. 20
Polymer 20
Cluster 4
Cluster 6
Method %
CO2 Immisc. 22.58 Method %
Air 12.90 N2 Misc. 42.86
Cluster 5 Water Flooding 12.90 N2 Inmisc. 21.43
CO2 Misc. 9.68
Method % Polymer 9.68
WAG N2 Misc. 14.29
Air 41.38 WAG HC Inmisc. 9.68 Cluster 6 Water Flooding 14.29
Steam 27.59 N2 Misc. 6.45 WAG HC Misc. 7.14
CO2 Immisc. 10.34 WAG HC Misc. 6.45
N2 Inmisc. 3.23
Polymer 8.62 Steam 3.23
WAG CO2 Inmisc. 5.17 WAG CO2 Misc. 3.23
CO2 Inmisc.
Water Flooding 5.17
N2 Inmisc. 1.72 CO2 Misc.
N2 Inmisc.
N2 Misc.
Polymer
Steam
Wag CO2 Inmisc.
Wag HC Inmisc.
Cluster 2
Wag CO2 Misc.
Method %
Water Flooding 38.46 Wag HC Misc.
WAG CO2 Misc. 13.46 Air
WAG HC Misc. 13.46
N2 Misc. 9.62 Water Flooding
CO2 Misc. 7.69 Reservoir A
N2 Inmisc. 7.69
Cluster 1 Reservoir B
Polymer 5.77
Method % Air 3.85
Water Flooding 29.17 Cluster 3
CO2 Misc. 20.83
Method %
Polymer 18.75
N2 Inmisc. 6.25
Water Flooding 48.28
Steam 6.25 Polymer 25.29
WAG HC Misc. 6.25 WAG CO2 Misc. 12.64
CO2 Immisc. 4.17 CO2 Misc. 10.34
WAG CO2 Misc. 4.17 N2 Inmisc. 1.15
N2 Misc. 2.08 WAG HC Misc. 1.15
WAG N2 Misc. 2.08 Steam 1.15
Fig. 5. Venezuelan Reservoir map in the international data base projection, located in clusters 4 and 5 (see dashed circles).
Cluster E
Method %
Cluster A Co2 Inmi 85.7
N2_Inmiscible 14.3
Method %
Aire 20.0
Co2 Inmi 20.0
Polymer 20.0
Cluster C
Wag-HCInmi 20.0 Method %
Water Flooding 20.0 Polymer 33.3
Water Flooding 16.7
Steam 16.7
Co2 Mis 16.7
Aire 16.7
CO2 Immisc.
CO2 Misc.
N2 Immisc.
N2 Misc.
Polymer
Steam
WAG-HC Immisc.
WAG-CO2 Misc.
Cluster B
Method % WAG-HC Misc.
Aire 28.6 air
Co2 Mis 14.3 Water flooding
N2_Miscible 14.3
Reservoir A
Wag-HCInmi 14.3
Wag-HCMisc 14.3
Cluster D
Method %
Water Flooding 14.3
Co2 Mis 16.7
N2_Miscible 16.7
Wag-Co2Misc 16.7
Wag-HCInmi 16.7
Wag-HCMisc 16.7
Water Flooding 16.7
Fig. 6. New Analysis of Cluster 4 for ReservoirA, clearly indicated by the dashed circle in cluster C.
SPE 78332 SELECTION OF EOR/IOR OPPORTUNITIES BASED ON MACHINE LEARNING 11
Cluster 1
Method %
Cluster 2 Steam 37.5
Method % Air 25
Air 52.94 Polymer 25
Cluster 3 Steam 47.06 Water Flooding 12.5
Method %
Air 70 Cluster 5
Steam 30 Method %
Air 44.44
Steam 22.22
CO2 Inmisc. 11.11
Polymer 11.11
Water Flooding 11.11
CO2 Inmisc.
N2 Inmisc.
Polymer
Steam
Cluster 4
Air
Method %
Water Flooding
CO2 Inmisc. 33.33
Air 22.22 Wag CO2 Inmisc.
Wag CO2 Inmisc. 22.22 Series20
Water Flooding 11.11
Polymer 11.11
Cluster 6
Method %
CO2 Inmisc. 40
N2 Inmisc. 20
Wag CO2 Inmisc. 20
Polymer 20
Fig.7. New Analysis of Cluster 5 for Reservoir B, indicated by the dashed brown ellipse. Several points indicate sensitivity analysis.