Auto-characterization NFR using least squares support vector regression

Note: This project was published in Arabian Journal of Geosciences.

Project description

Pressure transient response (PTR) of horizontal well in naturally fractured reservoirs (NFR) has a particular characteristic shape. This PTR is often used to estimate parameters of NFRs and detect their wellbore and boundary regimes. Interporosity flow coefficient (λ) and storativity ratio (ω) are two important parameters of the NFR that often estimated by matching process on the PTR. Since the matching techniques’ results are not often unique, in this study, the multi-output least squares support vector regression (MLS-SVR) is employed for simultaneous estimation of λ and ω. //

A databank of 500 PTRs for horizontal wells in naturally fractured reservoirs is generated by the finite element method, converted to the pressure derivative (PD) curves, and then used to develop and evaluate this auto-characterization paradigm. The predictive accuracy of the model is checked and validated by both smooth and noisy PTRs. The proposed model predicts ω and λ with overall absolute average relative deviations (AARD) of 0.186% and 3.754%, respectively. The correlation coefficients (R2) of 1 and 0.99992 are obtained for the prediction of ω and λ, respectively. The Leverage outlier detection technique justified that only less than 6% of the predictions are within the suspect region. This MLS-SVR model can be simply integrated with commercial pressure transient analysis (PTA) packages for accurate prediction of ω and λ even from the noisy PTRs.

Dataset

You can download dataset from Excel file.

The dataset includes 500 data within 6 clases that will be used for training ANN models.

Basic information

An overview of the files is provided below.

• src_python/ contains all python code files
• src_matlab/ contains all matlab code files
• Datageneration/ contains all matlab code files using for data generation.
• images/ contains all images used in the article.
• dataset.xlsx all generated pressure data that use as training and validation dataset.
• AJGS.pdf article in pdf format.
• LICENSE.txt is the MIT license.
• README.md contains basic information for the repository and detailed information for how to compile and reproduce the results.

Installation

you can clone and open directories using following in your terminal:

git clone https://github.com/acse-srm3018/NFRpredictionSVM

Documentation

The articles published and can be found here

References

• Vaferi B, Eslamloueyan R, Ghaffarian N (2016) Hydrocarbon reservoir model detection from pressure transient data using coupled artificial neural network-Wavelet transform approach. Appl Soft Comput J 47:63–75.
• Xu X, Chen L (2019) Projection of long-term care costs in China, 2020–2050: based on the Bayesian quantile regression method. Sustainability 11:3530.