Professional Documents
Culture Documents
2020_new technique to quantify the productivity of complex wells using artificial intelligence
2020_new technique to quantify the productivity of complex wells using artificial intelligence
2020_new technique to quantify the productivity of complex wells using artificial intelligence
Amjed Hassan, Mohamed Mahmoud, Abdulaziz Al-Majed, and Abdulazeez Abdulraheem, King Fahd University of
Petroleum & Minerals
This paper was prepared for presentation at the International Petroleum Technology Conference held in Dhahran, Saudi Arabia, 13 – 15 January 2020.
This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s).
Contents of the paper, as presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The
material, as presented, does not necessarily reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at
IPTC are subject to publication review by Sponsor Society Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial
purposes without the written consent of the International Petroleum Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of
not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented.
Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435.
Abstract
Nonconventional wells (NCWs) are applied to increase the well deliverability and access the difficult
formations. The nonconventional wells have been used to refer for the advanced wells such as highly
deviated, horizontal, fishbones or multilateral wells. These wells offer a great potential to maximize the
hydrocarbon recovery, however, it is difficult to predict their performance. In the literature, numerous
mathematical models were developed to predict the well-productivity. However, the available models have
been developed by employing one or more simplifying assumption(s), which may lead to over or under
estimate the hydrocarbon production. This paper presents an effective technique to estimate the productivity
for nonconventional wells.
In this work, artificial intelligence (AI) technique was utilized to determine the well-productivity for
wide range of conditions. The developed models can determine the well performance without introducing
the complexity associated with the numerical approaches. Artificial neural network was utilized to estimate
the hydrocarbon production for two types of nonconventional wells; fishbone multilateral and hydraulically
fractured horizontal wells. Reliable models are presented to quantify the performance of nonconventional
wells producing from heterogeneous and anisotropic formations. The developed models evaluate the
importance of reservoir properties and well configuration on the well deliverability. Total of 850 data sets
were utilized to construct the intelligence models and to validate the prediction performance. Moreover,
mathematical equations were extracted utilizing the optimized ANN models. The extracted correlations
showed acceptable prediction errors, the absolute error around 7.4% in average.
The novelty of this work is that effective models are proposed to quantify the productivity of
nonconventional wells. The proposed models can be utilized to refine commercial software to narrow down
the deviations between the actual measurements and simulation outputs. Also, this work can contribute to
enhance our understanding of the oil-field management by improving the prediction of well deliverability.
Consequently, this study can help in optimizing the well planning for complex wells such as hydraulically
fractured horizontal and fishbone multilateral wells.
2 IPTC-19706-Abstract
Key Words: nonconventional wells (NCWs), artificial intelligence (AI), well productivity, predictive
models
Introduction
Nonconventional (NCWs) or advanced wells are the highly deviated, horizontal or multilateral wells
Figure 1—Conventional and unconventional wells for producing hydrocarbon from underground reservoirs.
Predicting the well productivity is an essential factor in designing and completing the production well, as
well as selecting the artificial lift and stimulation processes (Guo et al., 2008). Several techniques have been
reported to estimate the well performance for unconventional wells. Different correlations and analytical
models were reported to determine the inflow performance relationship (IPR) (Borisov, 1984; Economides,
IPTC-19706-Abstract 3
1994; Furui et al., 2003; Yildiz, 2005). The most popular equations are Fetkovitch's and Vogel's correlations
(Furui et al., 2003; Ahmed et al., 2016). Recently, numerical simulators have been utilized to estimate
the well productivity, which resulted in significant reduction for the estimation error. Several reservoir
models were evaluated to simulate dual porosity system, triple porosity system, isotropic and anisotropic
permeability conditions (Ding et al., 2012; Filho et al., 2015). Moreover, computational fluid dynamics
coefficient analysis indicates that the permeability ratio has higher influence for the case of fishbone well
compared to the case of hydraulically fractured horizontal.
Table 1—Statistical parameters for the case of hydraulically fractured horizontal well.
Coefficient
93.757 33.537 70.533 142.764 45.878
of variation
Standard
211.275 2.499 693.099 685.121 1551.738 48712.516
Deviation
Coefficient
346.352 37.491 25.116 25.153 65.763 59.507
of variation
IPTC-19706-Abstract 5
Figure 3—The relative importance of production rate with input parameters used to train the ANN model, for the fishbone well.
respectively. The values of correlation coefficient and average error indicates the good performance of the
ANN model in determining the production rate for case of hydraulically fractured horizontal well. Figure
4 illustrates the performance of neural network model for predicating the productivity of hydraulically
fractured horizontal well. In general, the error profile for the validation data (green curve) is higher than
that of the training data (blue curve) while indicates that no model memorization was occurred for this ANN
Figure 4—Performance of ANN model for predicating the productivity of hydraulically fractured
horizontal well. The best epoch number is 6 with mean squared error (MSE) of 0.00132.
Figure 5—a cross plot between the actual and predicted flow rate for the hydraulically fractured
horizontal well. The average absolute error is 6.82% and the R-value is 0.995 training data.
IPTC-19706-Abstract 7
Figure 7—Error profiles for the training and testing data during the predication of the
fishbone productivity. The best epoch number is 4 with mean squared error (MSE) of 0.0036.
8 IPTC-19706-Abstract
Figure 9—Cross plot between the actual and predicted flow rate for the fishbone multilateral
well. For the testing data, the average absolute error and the R-value are 6.36% and 0.995.
Comparison Analysis
Figure 10 shows the predicted and actual flow rate for fishbone and hydraulically fractured horizontal wells.
The flow rate for both wells showed a good aligned around the 45° line indicating the effective prediction
performance for the developed ANN models. Also, it was observed than the match between the actual and
predicted values become better to higher flow rates. Furthermore, the relationship between the prediction
performance and the flowing bottomhole pressure was studied. Figure 11 shows the error profiles against
IPTC-19706-Abstract 9
the dimensionless pressure. Increasing the flowing pressure can lead to increase the prediction error for both
wells. However, at same pressure, higher error was observed for the prediction of hydraulically fractured
horizontal well compared to the fishbone well.
Figure 11—Average percentage error against the dimensionless pressure for the fishbone and
hydraulically fractured horizontal wells, Higher error was observed at higher flowing pressure.
10 IPTC-19706-Abstract
Empirical Correlations
In order to convert the ANN model from black box into a white box, empirical correlations were extracted
from the optimized ANN models. The extracted correlations can provide a direct method for determining
the well productivity. The weights (w) and biases (b) of the neural network model were utilized to derive
the empirical correlations. For the hydraulically fractured horizontal well, the empirical correlation is given
(1)
(2)
For the fishbone multilateral well, the empirical equation is given as;
(3)
Where, QD is the dimensionless flow rate, PD is the dimensionless pressure, n is the number of fractures,
CfD is the dimensionless fracture conductivity, L is the lateral length and Kv/Kh is the permeability ratio.
The values of weights and biases of hidden and output layers are summarized in Table 3 and 4 for the
hydraulically fractured horizontal and fishbone multilateral wells, respectively.
Table 3—The values of weights and biases for hydraulically fractured horizontal well.
Table 4—The values of weights and biases for the fishbone multilateral well.
Conclusions
This paper presents effective models to quantify the productivity of complex wells using artificial
intelligence technique. The production rates for hydraulically fractured horizontal well and fishbone
multilateral well were determined using artificial neural network (ANN). The reservoir properties and the
well configurations were used to estimate the well performances. The developed models can determine
the well production without introducing the complexity associated with the numerical approaches. Average
absolute percentage error (AAPE) and correlation coefficient were used to select the best ANN models.
The developed models can determine the flow rate for hydraulically fractured horizontal and fishbone
multilateral wells with average errors of 8.39% and 6.36%, respectively, and correlation coefficients higher
than 0.99 for both wells. Moreover, empirical correlations were extracted from the optimized ANN models.
The extracted correlations can provide a direct and simple approach to calculate the well productivity.
The proposed equations can be inserted into the commercial software to allow better prediction for the
productivity of complex well such as hydraulically fractured horizontal and fishbone multilateral wells.
Acknowledgement
Authors would like to acknowledge College of Petroleum and Geosciences (CPG), at King Fahd University
of Petroleum and Minerals (KFUPM) for the technical supports and permission to publish this paper.
References
Abdalla, M., Hassan, A., Abdulraheem, A., Elkatatny, S. and Mohamed, A., 2018, January. New Technique to Evaluate the
Performance of Hydraulically Fractured Horizontal Wells. In SPE/IADC Middle East Drilling Technology Conference
and Exhibition. Society of Petroleum Engineers.
Ahmed, M. E., Alnuaim, S. and Abdulazeem, A. New Algorithm to Quantify Productivity of Fishbone Type Multilateral
Gas Well. SPE-181888-MS, presented at the SPE Annual Technical Conference and Exhibition held in Dubai, UAE,
2016, 26-28 September.
AlAjmi, M. D., Alarifi, S. A., Mahsoon, A. H. Improving Multiphase Choke Performance Prediction and Well Production
Test Validation Using Artificial Intelligence: A New Milestone. SPE-173394-MS, presented at the SPE Digital Energy
Conference and Exhibition, held in The Woodlands, Texas, USA, 2015, 3-5 March.
Alarifi, S. A., AlNuaim, S., Abdulraheem, A. Productivity Index Prediction for Oil Horizontal Wells Using Different
Artificial Intelligence Techniques. SPE-172729-MS, presented at the SPE Middle East Oil & Gas Show and
Conference, held in Manama, Bahrain, 2015, 8-11 March.
12 IPTC-19706-Abstract
Al-Mashhad A. S., Al-Arifi S. A., Al-Kadem M. S., Al-Dabbous M. S., Buhulaigah A. Multilateral Wells Evaluation
Utilizing Artificial Intelligence. SPE-183508-MS, presented at the Abu Dhabi International Petroleum Exhibition and
Conference held in Abu Dhabi, UAE, 2016, 7–10 November.
Borisov, Ju. P. Oil Production using Horizontal and multiple deviation wells. J. strauss, Trans. And S.D Joshi, Ed., Phillips
Petroleum Co. Bartlesville, 1984.
Bosworth, X., El-Sayed, H.S., Ismail, G., Ohmer, H., Stracke, M., West, C., Retnanto, A. Key Issues in Multilateral