0
Research Papers: Petroleum Engineering

Practical Data Mining and Artificial Neural Network Modeling for Steam-Assisted Gravity Drainage Production Analysis

[+] Author and Article Information
Zhiwei Ma

School of Mining and Petroleum Engineering,
Department of Civil and Environmental Engineering,
University of Alberta,
Edmonton, AB T6H 1G9, Canada
e-mail: zma2@ualberta.ca

Juliana Y. Leung

School of Mining and Petroleum Engineering,
Department of Civil and Environmental Engineering,
University of Alberta,
Edmonton, AB T6H 1G9, Canada
e-mail: juliana2@ualberta.ca

Stefan Zanon

Nexen Energy ULC,
801-7th Avenue S.W.,
Calgary, AB T2P 3P7, Canada
e-mail: stefan.zanon@nexencnoocltd.com

1Corresponding author.

Contributed by the Petroleum Division of ASME for publication in the JOURNAL OF ENERGY RESOURCES TECHNOLOGY. Manuscript received August 27, 2016; final manuscript received December 30, 2016; published online February 8, 2017. Assoc. Editor: Daoyong (Tony) Yang.

J. Energy Resour. Technol 139(3), 032909 (Feb 08, 2017) (10 pages) Paper No: JERT-16-1347; doi: 10.1115/1.4035751 History: Received August 27, 2016; Revised December 30, 2016

Production forecast of steam-assisted gravity drainage (SAGD) in heterogeneous reservoir is important for reservoir management and optimization of development strategies for oil sand operations. In this work, artificial intelligence (AI) approaches are employed as a complementary tool for production forecast and pattern recognition of highly nonlinear relationships between system variables. Field data from more than 2000 wells are extracted from various publicly available sources. It consists of petrophysical log measurements, production and injection profiles. Analysis of a raw dataset of this magnitude for SAGD reservoirs has not been published in the literature, although a previous study presented a much smaller dataset. This paper attempts to discuss and address a number of the challenges encountered. After a detailed exploratory data analysis, a refined dataset encompassing ten different SAGD operating fields with 153 complete well pairs is assembled for prediction model construction. Artificial neural network (ANN) is employed to facilitate the production performance analysis by calibrating the reservoir heterogeneities and operating constraints with production performance. The impact of extrapolation of the petrophysical parameters from the nearby vertical well is assessed. As a result, an additional input attribute is introduced to capture the uncertainty in extrapolation, while a new output attribute is incorporated as a quantitative measure of the process efficiency. Data-mining algorithms including principal components analysis (PCA) and cluster analysis are applied to improve prediction quality and model robustness by removing data correlation and by identifying internal structures among the dataset, which are novel extensions to the previous SAGD analysis study. Finally, statistical analysis is conducted to study the uncertainties in the final ANN predictions. The modeling results are demonstrated to be both reliable and acceptable. This paper demonstrates the combination of AI-based approaches and data-mining analysis can facilitate practical field data analysis, which is often prone to uncertainties, errors, biases, and noises, with high reliability and feasibility. Considering that many important system variables are typically unavailable in the public domain and, hence, are missing in the dataset, this work illustrates how practical AI approaches can be tailored to construct models capable of predicting SAGD recovery performance from only log-derived and operational variables. It also demonstrates the potential of AI models in assisting conventional SAGD analysis.

Copyright © 2017 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

Typical SAGD project: (a) schematic of a single well pair in 3D and (b) schematic of a well pad consisting of five well pairs in top view

Grahic Jump Location
Fig. 2

Neural network architecture with only one hidden layer

Grahic Jump Location
Fig. 3

Flowchart of the adopted analysis workflow

Grahic Jump Location
Fig. 4

Correlation plot of the original dataset: large positive value indicates a strong positive correlation between the two parameters; low negative value indicates a strong negative correlation between the two parameters

Grahic Jump Location
Fig. 5

Input variable importance plot

Grahic Jump Location
Fig. 6

Principal component analysis: (a) variance plot and (b) bi-plot: visualization of the orthonormal principal component coefficients for each variable with respect to the first two principal components. The 153 data samples are denoted by dots.

Grahic Jump Location
Fig. 7

Scatter plot between the first principal score (PS 1) and the second principal score (PS 2) for all ten producing fields: smaller markers—cluster 1; larger markers—cluster 2

Grahic Jump Location
Fig. 8

Cluster analysis results of the ten producing fields

Grahic Jump Location
Fig. 9

Comparison of histograms of the ten original input variables from cluster 1 and cluster 2

Grahic Jump Location
Fig. 10

Cross plots of TISOR and COP from ANN prediction and target TISOR and COP from field data by using manual grouping based on the input parameter d: group 1—data samples with d larger than the median of 22 m (top); group 2—data samples with d less than the median of 22 m (bottom)

Grahic Jump Location
Fig. 11

Cross plots of TISOR and COP from ANN prediction and target TISOR and COP from field data following k-mean cluster analysis: top—cluster 1; bottom—cluster 2

Grahic Jump Location
Fig. 12

Residual error analysis for the two outputs: TISOR (a) and COP (b). Top, middle and bottom row represent the residual error, relative error, and distribution of corresponding output parameter, respectively.

Grahic Jump Location
Fig. 13

Cross plots of TISOR and COP from ANN prediction and target TISOR and COP from field data without PCA: top—cluster 1; bottom—cluster 2

Grahic Jump Location
Fig. 14

Cross plots of TISOR and COP from ANN prediction and target TISOR and COP from field data: switching the testing datasets between two clusters while keeping the training datasets unchanged: top—cluster 1; bottom—cluster 2

Grahic Jump Location
Fig. 15

Cross plots of TISOR and COP from ANN prediction and target TISOR and COP from field data: without cluster analysis

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In