Professional Documents
Culture Documents
Hybrid computational models for the characterization of oil and gas reservoirs
Tarek Helmy *,1, Anifowose Fatai, Kanaan Faisal
Information and Computer Science Department, College of Computer Science and Engineering, King Fahd University of Petroleum and Minerals, Mail Box. 413,
Dhahran 31261, Saudi Arabia
a r t i c l e i n f o a b s t r a c t
Keywords: The process of combining multiple computational intelligence techniques to build a single hybrid model
Hybrid computational intelligence has become increasingly popular. As reported in the literature, the performance indices of these hybrid
Petroleum reservoir characteristics models have proved to be better than the individual components when used alone. Hybrid models are
Fuzzy Logic extremely useful for reservoir characterization in petroleum engineering, which requires high-accuracy
Support Vector Machines
predictions for efficient exploration and management of oil and gas resources.
Functional Networks
In this paper, we have utilized the capabilities of data mining and computational intelligence in the
prediction of porosity and permeability, two important petroleum reservoir characteristics, based on
the hybridization of Fuzzy Logic, Support Vector Machines, and Functional Networks, using several
real-life well-logs. Two hybrid models have been built. In both, Functional Networks were used to select
the best of the predictor variables for training directly from input data by using its functional approxima-
tion capability with least square fitting algorithm. In the first model (FFS), the selected predictor variables
were passed to Type-2 Fuzzy Logic System to handle uncertainties and extract inference rules, while Sup-
port Vector Machines made the final predictions. In the second, the selected predictor variables were
passed to Support Vector Machines for training by transforming them to a higher dimensional space,
and then to Type-2 Fuzzy Logic to handle uncertainties, extract inference rules and make final predic-
tions.
The simulation results show that the hybrid models perform better than the individual techniques
when used alone for the same datasets with their higher correlation coefficients. In terms of execution
time, the hybrid models took less time for both training and testing than the Type-2 Fuzzy Logic, but
more time than Functional Networks and Support Vector Machines. This could be the price for having
a better and more robust model. The hybrid models also performed better than a combination of two
of the individual components, Type-2 Fuzzy Logic and Support Vector Machines, in terms of higher cor-
relation coefficients as well as lower execution times. This is due to the effective role of Functional Net-
works, as a best-variable selector in the hybrid models.
Ó 2010 Elsevier Ltd. All rights reserved.
0957-4174/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eswa.2010.01.021
5354 T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363
integrity, porosity and permeability. CI covers many existing sediment. Permeability is a key parameter associated with the
branches of science viz. Artificial Neural Networks, Fuzzy Systems, characterization of any hydrocarbon reservoir. In fact, it is not pos-
Evolutionary Computing and Hybrid Systems. It also embraces sible to have accurate solutions to many petroleum engineering
techniques that use Swarm Intelligence, Fractals and Chaos Theory, problems without having accurate permeability value (K.S. Univer-
Artificial Immune Systems, Wavelets, etc. (He, 2008). sity, 2007).
The combination of two or more Computational Intelligence El Ouahed et al. (2005) produced 2-D fracture intensity and
schemes as a single model is called Hybrid Computational Intelli- fracture network maps in a large block of field using Artificial Neu-
gence (HCI) and is becoming increasingly popular. This increased ral Network and FL. The results showed that the proposed ap-
popularity lies in the extensive success of hybrid systems in many proach is a practical methodology to map the fracture network.
real-world complex problems (Giovanni & Loia, 2005). A key prere- Ali and Chawathe (2000) used neural networks to predict perme-
quisite for the merging of technologies is the existence of a ‘‘com- ability from petrographic data while using Fuzzy Logic to screen
mon denominator” to build upon (Andreas, Symeonidis, & Pericles and rank the predictor variables with respect to the target variable.
Mitkas, 2005). In our case, part of the ‘‘common denominator” for The result demonstrated the generalizing capability of the neural
Fuzzy Logic (FL), Support Vector Machines (SVM) and Functional network. A similar study was done by Saemi, Ahmadi, and Varjani
Networks (FN) is the inference procedures they deploy and their (2007) when they proposed a new method for the auto-design of
excellent predictive capabilities. We attempt to combine the indi- neural networks based on genetic algorithm (GA).
vidual capabilities of FL, SVM and FN in an HCI scheme, to predict The traditional FL, now referred to as Type-1 FL has featured in a
two characteristics of oil and gas reservoirs, namely porosity and number of research efforts, especially in reservoir characterization.
permeability, with better performance indices. Type-2 FL has also featured in many recently published articles in
Our motivations for this work include the quest for higher per- various fields and especially in reservoir properties modeling (El
formance accuracy in the prediction of oil and gas characteristics, Ouahed et al., 2005). One of the earliest references to the applica-
the recently increasing popularity of hybrid intelligent systems, tion of FL in the petroleum industry was by Fang and Chen (1997)
the reported success of these systems in many real-world complex who presented a fuzzy modeling for predicting porosity and per-
problems, the need to complement the weaknesses of one algo- meability from the compositional and textural characteristics of
rithm with the advantages of the others and hence to combine sandstones. They found that fuzzy modeling is assumption-free,
the cooperative and competitive characteristics of the individual tolerant of outliers, and capable of making both linguistic and nu-
techniques and the existing theoretic and experimental justifica- meric predictions based on qualitative and quantitative data.
tions (Munakata, 2008; Nong, 2003; Peddabachigari, Abrahamb, SVMs have been used extensively in many areas, including oil
Grosan, & Thomas, 2007; Phillips-Wren, Ichalkaranje, & Jain, and gas (Jian & Wenfen, 2006), with very promising results. Tabo-
2008; Sun et al., 2007; Triantaphyllou & Felici, 2006; Wang & Fu, ada, Matías, Ordóñez, and García (2007) used different kinds of
2005) that hybrids produce more accurate results than the individ- SVMs: SVM classification (multi-class one-against-all), ordinal
ual techniques used separately. SVM and SVM regression, and they found that the SVMs are per-
The rest of this paper is organized as follows: Section 2 reviews fectly comparable to kriging (a statistical model) and have better
the literature on how Artificial Intelligence (AI) can be employed in control of outliers.
Petroleum Engineering. Section 3 presents a brief overview of Fuz- Like the other AI techniques described in previous sections,
zy Logic, Support Vector Machines, Functional Networks and Hy- FunNets have also featured in a number of research studies. Castil-
brid Systems. Section 4 explains the Well log data and tools, lo (1998) and Castillo, Hadi, and Lacruz (2001a) gave a comprehen-
along with the methodology used in this work. Section 5 discusses sive demonstration of the application of FunNets in Statistics and
the criteria of quality measurement used to evaluate the validity in Engineering.
this work while conclusion, with a detailed plan of future work is Most hybrids found in the literature usually contain Neural Net-
described in Section 6. works fused with one other technique due to its wide use in the
computational intelligence. Other hybrid systems include different
combinations of several techniques such as SVM, FL, Decision
2. Literature survey Trees, Extreme Learning Machines, Genetic Algorithm, Hidden
Markov Model and Radial-basis Functions, as applied in various
The application of the capabilities of Artificial Intelligence (AI) fields including reservoir characterization (Chikhi, 2006; Chikhi &
has been widely appreciated in petroleum engineering, as well as Batouche, 2004).
in other fields. Some of the areas of petroleum technology in which A fuzzy linear programming Support Vector Machine (LP-SVM)
AI has been used with success include seismic pattern recognition, was used for multiclass classification problems by Shigeo (2004),
porosity and permeability predictions, identification of sandstone demonstrating the superiority of the hybrid over the conventional
lithofacies, drill bit diagnosis, and analysis and improvement of SVMs. Deny, Wilkinson, Yu, and Ramon (2005) developed a hybrid
gas well production (Ali, 1994; El Ouahed, Tiab, & Mazouzi, genetic programming and fuzzy/neural network inference system
2005; Goda, Maier, & Behrenbruch, 2007). Porosity and permeabil- to estimate the permeability of reservoirs. When compared with
ity measurements are frequently made on plugs extracted from the contemporary estimation approaches, the hybrid yielded more
core of wells drilled for oil and gas exploration. The data are valu- consistent and robust estimated results. Chikhi and Batouche
able for linking permeability to porosity, and these both serve as (2004) combined a neural method with radial-basis functions to
standard indicators of reservoir quality in the oil and gas industry. enhance the classification of lithofacies of certain wells. This was
Porosity is the percentage of voids and open space in a rock or followed by a similar hybridization effort by Chikhi (2006), who
sedimentary deposit. The greater the porosity of a rock, the greater combined Neural Networks with Hidden Markov Models to obtain
its ability to hold water and other earth materials, such as oil the lithological identification of the same wells. The results ob-
(Excellence in Educational Development, 2007). Porosity is very tained by the hybrid are close to those obtained by the fuzzy Adap-
important in evaluating the potential volume of hydrocarbons it tive Resonance Theory (ART) approach applied to the same
may contain. Permeability is the ease with which fluid is transmit- borehole with the same well-logs. A hybridization of SVMs and
ted through a rock’s pore space. Although a rock may be very por- the Interval Type-2 FL System (FLS) was performed by Chen, Li,
ous, it is not necessarily very permeable. Permeability is a measure Harrison, and Zhang (2007) to better handle uncertainties existing
of how interconnected the individual pore spaces are in a rock or in real classification data.
T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363 5355
Regression in Primal
Y
Training
Dataset
3.2. Support Vector Machines An approach resulting from the combination of two or more ap-
proaches is called a hybrid. It has also been defined as an approach
Support Vector Machines are a set of related supervised learn- that combines different theoretical backgrounds and algorithms
ing methods used for classification and regression. They belong such as data mining and soft computing methodologies. The main
to a family of generalized Linear Classifiers. They can also be con- idea behind hybridization is to complement the weaknesses of one
sidered as a special case of Tikhonov Regularization. SVMs map in- technique with the strength of other techniques. Since no single
put vectors to a higher dimensional space where a maximal technique is good for everything and in all situations, there is a
separating hyperplane is constructed (Burges, 1998; Littman, need to combine the individual capabilities of each technique to
2003). This is shown in Fig. 3. obtain a more versatile and robust technique.
5356 T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363
form in North America (site 1) and the three for permeability from
a drilling site in the Middle East (site 2). The datasets from site 1
have six predictor variables for porosity, while the dataset from
site 2 has eight predictor variables for permeability. These are
shown in Tables 1 and 2.
Table 3
Strengths and weaknesses of FL, SVM and FN.
Handling Transformation to
Uncertainties from higher dimensional
Least Square
best-variable data space for regression
Fitting
Handling
Least Square Uncertainties from
Fitting best-variable data
Table 5
Result of the porosity prediction site 1, well 1.
Table 6
Result of the porosity prediction site 1, well 2.
Table 7
Result of the porosity prediction site 1, well 3.
5.5. Criteria for performance evaluation The RMSE is one of the most commonly used measures of suc-
cess for numeric prediction, computed by taking the square root of
In order to establish a valid and reasonable evaluation of this the average of the squared differences between each predicted va-
work, we have used Correlation Coefficient (CC), Root Mean- lue xn and its corresponding actual value yn. It gives the error value
Squared Error (RMSE), and Execution Time (ET) as criteria for mea- the same dimensionality as the actual and predicted values
suring the performance. The CC measures the statistical correlation
between the predicted and actual values. A value of ‘‘1” means per- qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðx1 y1 Þ2 þðx2 y2 Þ2 þþðxn yn Þ2
fect statistical correlation and a ‘‘0” means there is no correlation at The formula is : n :
all. where n is the size of data
P
ðx x0 Þðy y0 Þ
The formula is : qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P P
ðx xÞ2 ðy y0 Þ2 The ET is simply the total time taken for a technique to run from the
beginning to the end, and it is computed as: T 2 T 1 , where T2 is the
where x and y are the actual and predicted values, while x0 and y0 are CPU time at the end of the run and T1 is the CPU time at the begin-
the mean of the actual and predicted values. ning of the run.
T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363 5359
Fig. 9. Execution time comparison for well 3 porosity training and testing.
Table 8
Result of the permeability prediction for site 2, well 1.
Table 9
Result of the permeability prediction for site 2, well 2.
6. Experimental results Tables 8–10 present the result of the prediction of Permeability
for the three wells in site 2.
We implemented and validated the individual techniques as These are also summarized graphically in the plots shown in
well as the two hybrid models in the prediction of porosity Figs. 10 and 11 for both training and testing.
and permeability by using the training and testing data In order to appreciate the role performed by the FN block in the
described in the previous section. Several iterations were made, FFS and FSF hybrid systems, another set of experiments was per-
and the average of results was taken. Tables 5–7 present the formed by using only two of the components namely: Type-2 Fuz-
result of the prediction of porosity for the three wells in zy Logic and SVM.
site 1. The results of the simulations for the three Porosity and three
These are summarized graphically in Figs. 7–9. Permeability wells are shown in Tables 11 and 12, respectively.
5360 T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363
Table 10
Result of the permeability prediction for site 2, well 3
Fig. 10. Correlation coefficients comparisons for permeability training and testing.
Fig. 11. Execution time comparisons for porosity training and testing.
Table 11
Results of the simulations for the three porosity wells.
Comparisons of the performance of the Type-2-SVM above with In terms of execution time, the results showed that Functional
the FFS and FSF Hybrids are shown in Figs. 12–15. Networks are the fastest in terms of both training and testing, fol-
lowed by SVM. Type-2 Fuzzy Logic took the most time for both
6.1. Discussion of results training and testing, due to its complexity as described in Sec-
tion 3.1 and Table 3. The hybrid models proved to be faster than
In the prediction of porosity and permeability, it is clear from the Fuzzy Logic component, but not for SVM and FN. This is the
the results that the hybrid models performed better than, or com- price for obtaining better models in terms reliability and
petitively equal to, the three individual techniques used separately, robustness.
in terms of their correlation coefficient. A special characteristic was When compared to the Type-2-SVM Hybrid, the FFS and FSF Hy-
observed in the result of site 1 well 3 for Porosity, where SVM dem- brids also proved to be better in terms of both correlation coeffi-
onstrated its ability to withstand a shortage of training data but FL cient and execution time. The better performance in correlation
demonstrated otherwise. Still, the hybrid models performed better coefficient can be attributed to the role of Functional Networks
than the individual components due to the cooperative spirit that in the FFS and FSF hybrids. The Functional Networks block serves
was been built into them. as a best-variable selector, which extracts from the input variables
T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363 5361
Table 12
Results of the simulations for the three permeability wells.
Fig. 12. Correlation coefficient comparisons for porosity training and testing.
Fig. 13. Execution time comparisons for porosity training and testing.
Fig. 14. Correlation coefficient comparisons for permeability training and testing.
only those variables that are most relevant to the prediction The better performance of the FFS and FSF hybrids in terms of
system. execution times is also due to the above reason. In the process of
5362 T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363
Fig. 15. Execution time comparisons for permeability training and testing.
best-variable selection by the Functional Networks block, the Andreas, L., Symeonidis, L., & Pericles Mitkas, A. (2005). Agent intelligence through
data mining, Multiagent systems, artificial societies, and simulated organizations
dimensionality of the input variable that goes to the next block is
series, Vol. 14. International Book Series.
reduced. This works in favor of the Type-2 FLS that does not work Burges, C. J. C. (1998). A tutorial on support vector machines for pattern recognition.
well with an input data of very high dimensionality. Data Mining and Knowledge Discovery, 2, 121–167.
Castillo, E. (1998). Functional networks. Neural Processing Letters, 7, 151–159.
Castillo, E., Gutiérrez, J. M., Hadi, A., & Lacruz, B. (2001). Some applications
7. Conclusion of functional networks in statistics and engineering. Technometrics 43,
10–24.
Castillo, E., Hadi, A., & Lacruz, B. (2001). Optimal transformations in multiple linear
Two conceptual design frameworks for the hybridization of Fuz- regression using functional networks. In International work-conference on
zy Logic, Support Vector Machines and Functional Networks have artificial and natural neural networks. Lecture Notes in Computer Science (Vol.
been implemented and presented. They were tested by using two 2084, pp. 316–324).
Chen, X., Li, Y., Harrison, R., & Zhang, Y. Q. (2007). Type-2 fuzzy logic based classifier
well-logs containing six porosity and permeability datasets. The fusion for support vector machines. Applied Soft Computing Journal. doi:10.1016/
results showed that the hybrids performed better than the individ- j.asoc.2007.02.019.
ual techniques used separately for all the datasets in the prediction Chikhi, S. (2006). A fuzzy ART versus hybrid NN-HMM methods for lithology
identification in the Triassic Province. IEEE Transactions, 7803-9521-2/06.
of porosity and permeability. With any novel data, the hybrid will
Chikhi, S., & Batouche, M. (2004). Probabilistic neural method combined with
be able to select the best model to use, extract inference rules, and radial-bias functions applied to reservoir characterization in the Algerian
handle uncertainties that might be present, directly from the data, Triassic Province. Journal of Geophysics and Engineering, 1, 134–142.
Deny, X., Wilkinson, D., Yu, T., & Ramon, S. (2005). California, permeability
and perform the required prediction. The hybrid models are also
estimation using a hybrid genetic programming and fuzzy/neural inference
seen to be very reliable, robust and effective, given the reported approach. In 2005 Society of petroleum engineers annual technical conference and
and observed good performance of the individual components. exhibition held in Dallas, TX, USA, 9–12 October, 2005.
Despite the good performance of the hybrid, we could not fully Duch, W., Adamczak, R., & Jankowski, N. (1997). Initialization and optimization of
multilayered perceptrons. In: Third conference on neural networks and their
validate its performance due to lack of data that contain uncertain applications, Kule, Poland, 1997.
measurements. The data we used had no such uncertainty. Conse- El Ouahed, A. K., Tiab, D., & Mazouzi, A. (2005). Application of artificial intelligence
quently, the design, implementation and validation of the models to characterize naturally fractured zones in Hassi Messaoud Oil Field, Algeria.
Journal of Petroleum Science and Engineering, 49, 122–141.
were based on well-logs and not on expert knowledge. Also, the S. Excellence in Educational Development, Science Lab Project (2007). <http://
Type-2 FLS has some limitations, and our framework was built www.seed.slb.com/en/scictr/lab/porosity/index.htm>.
on these limitations. Fang, J. H., & Chen, H. C. (1997). Fuzzy modeling and the prediction of porosity and
permeability from the compositional and textural attributes of sandstone.
Motivated by the success of this work, part of our future work Journal of Petroleum Geology, 20(2), 185–204.
will be directed towards modifying this hybrid model in order to Giovanni, A., & Vincenzo, L. (2005). Using FML and Fuzzy technology in adaptive
solve classification and pattern recognition problems in the oil ambient intelligence environments. International Journal of Computational
Intelligence Research, 1(2), 171–182., ISSN: 0973-1873. <http://www.ijcir.info>.
and gas industry such as Lithofacie and History Matching. More
Goda, H. M., Maier, & Behrenbruch, H. R. (2007). Use of artificial intelligence
experiments will be done on these hybrids if data is available with techniques for predicting irreducible water saturation – Australian
uncertain numerical measurements, and the frameworks will be Hydrocarbons Basins. In Society of Petroleum Engineers Asia Pacific Oil & Gas
Conference and Exhibition, Jakarta, Indonesia, 2007.
validated by using expert data instead of well-logs.
Guojie, L. (2004). Radial basis function neural network for speaker verification. A
Master of Engineering Thesis. The Nanyang Technological University.
Acknowledgments He, J. (2008). Computational intelligence. Research interests, <http://
www.cs.bham.ac.uk/~jxh/hejunrs.html>.
Hosmer, D., & Lemeshow, S. (2000). Applied logistic regression (p. 392). John Wiley
We would like to thank both KACST and King Fahd University of and Sons.
Petroleum and Minerals for providing the computing facilities and Jian, H., & Wenfen, H. (2006). Novel approach to predict potentiality of enhanced oil
support. Special thanks go to Mr. David Birkett for his help in proof recovery. In Society of petroleum engineers intelligent energy conference and
exhibition, Amsterdam, The Netherlands, 2006.
reading of the paper. K.S. University. (2007). Electronic document Archive, docs.ksu.edu.sa/DOC/
Articles49/Article490822.doc.
References Karnik, N. N., & Mendel, J. M. (1999). Type-2 fuzzy logic systems. IEEE Transactions
on Fuzzy Systems, 7(6), 5.
Eitel Lauría, J. M., & Peter Duchessi, J. (2006). A Bayesian belief network for it
Ali, J. K. (1994). Neural networks: A new tool for the petroleum industry? In
implementation decision support. In Decision support systems and international
European petroleum computer conference, Aberdeen, UK, March 1994.
federation for information processing international conference on decision support
Ali, M., & Chawathe, A. (2000). Using artificial intelligence to predict permeability
systems, 2006.
from petrographic data. Computers & Geosciences, 26, 915–925.
T. Helmy et al. / Expert Systems with Applications 37 (2010) 5353–5363 5363
Li Sun, Z., Au, K., & Choi, T. (2007). A neuro-fuzzy inference system through Saemi, M., Ahmadi, M., & Varjani, A. Y. (2007). Design of neural networks using
integration of fuzzy logic and extreme learning machines. IEEE Transactions on genetic algorithm for the permeability estimation of the reservoir. Journal of
Systems, Man, and Cybernetics – Part B: Cybernetics, 37(5), 1321–1331. Petroleum Science and Engineering, 59, 97–105.
Lim, J. (2005). Reservoir properties determination using fuzzy logic and neural Salah, A., Rahman, S., & Nath, K. (2005). An enhancement of K-nearest neighbor
networks from well data in offshore Korea. Journal of Petroleum Science and classification using genetic algorithm. In Midwest instruction and computing
Engineering, 49, 182–192. symposium.
Littman, W. T. A. (2003). Introduction to support vector machines. Machine learning Shigeo, A. (2004). Fuzzy LP-SVMs for multiclass problems. In Proceedings of the
course 536. In Tom Mitchell, Text book: Machine learning, McGraw Hill, 1997. European symposium on artificial neural networks, Bruges, Belgium, pp. 429–434.
Department of Computer Science, Rutgers University, The State University of Taboada, J., Matías, J. M., Ordóñez, C., & García, P. J. (2007). Creating a quality map of
New Jersey, USA. <http://www.cs.rutgers.edu/~mlittman/courses/ml03/>. a slate deposit using support vector machines. Journal of Computational and
Mendel, J. M. (2003). Type-2 fuzzy sets: Some questions and answers. IEEE Applied Mathematics, 204, 84–94.
Connections, Newsletter of the IEEE Neural Networks Society, 1, 10–13. Triantaphyllou, E., & Felici, G. (2006). Data mining and knowledge discovery
Munakata, T. (2008). Fundamentals of the new artificial intelligence neural, approaches based on rule induction techniques. Springer Science-i-Business
evolutionary, fuzzy and more (2nd ed.). Springer Science + Business Media. Media. 48.
Nong, Y. (2003). The handbook of data mining pp. 561, 617. Lawrence Erlbaum Wang, L., & Fu, X. (2005). Data mining with computational intelligence (pp. 229).
Associates. Berlin, Heidelberg: Springer-Verlag.
Peddabachigari, S., Abrahamb, A., Grosan, C., & Thomas, J. (2007). Modeling Zarandi, M. H. F., Rezaee, B., Turksen, I. B., & Neshat, E. (2007). A Type-2 fuzzy rule-
intrusion detection system using hybrid intelligent systems. Journal of based expert system model for stock price analysis. Expert Systems with
Network and Computer Applications, 30, 114–132. Applications, 9–34. doi:10.1016/j.eswa.2007.09.034.
Phillips-Wren, G., Ichalkaranje, N., & Jain, C. C. (2008). Intelligent decision making: An
AI-based approach, studies in computational intelligence (Vol. 97, pp. 88, 102,
355). Berlin, Heidelberg: Springer-Verlag.