Journal
JOURNAL OF APPLIED GEOPHYSICS
Volume 89, Issue -, Pages 96-107Publisher
ELSEVIER
DOI: 10.1016/j.jappgeo.2012.11.010
Keywords
Compressional wave velocity; Adaptive neuro fuzzy inference system; Multiple regression; Carbonate reservoirs
Funding
- University Teknologi Malaysia
Ask authors/readers for more resources
Compressional-wave (V-p) data are key information for estimation of rock physical properties and formation evaluation in hydrocarbon reservoirs. However, the absence of V-p will significantly delay the application of specific risk-assessment approaches for reservoir exploration and development procedures. Since V-p is affected by several factors such as lithology, porosity, density, and etc., it is difficult to model their non-linear relationships using conventional approaches. In addition, currently available techniques are not efficient for V-p prediction, especially in carbonates. There is a growing interest in incorporating advanced technologies for an accurate prediction of lacking data in wells. The objectives of this study, therefore, are to analyze and predict V-p as a function of some conventional well logs by two approaches; Adaptive Neuro-Fuzzy Inference System (ANFIS) and Multiple Linear Regression (MLR). Also, the significant impact of selected input parameters on response variable will be investigated. A total of 2156 data points from a giant Middle Eastern carbonate reservoir, derived from conventional well logs and Dipole Sonic Imager (DSI) log were utilized in this study. The quality of the prediction was quantified in terms of the mean squared error (MSE), correlation coefficient (R-square), and prediction efficiency error (PEE). Results show that the ANFIS outperforms MLR with MSE of 0.0552, R-square of 0.964, and PEE of 2%. It is posited that porosity has a significant impact in predicting V-p in the investigated carbonate reservoir. (c) 2012 Elsevier B.V. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available