@inproceedings{scholars4838,
            note = {cited By 0; Conference of 3rd International Geobaikal Conference 2014: Exploration and Field Development in East Siberia, GEOBAIKAL 2014 ; Conference Date: 18 August 2014 Through 22 August 2014; Conference Code:114659},
            year = {2014},
           title = {Integrated analysis of well logs and seismic data for reservoir characterization to estimate hydrocarbon},
       publisher = {European Association of Geoscientists and Engineers, EAGE},
         journal = {GEOBAIKAL 2014 - 3rd International Geobaikal Conference 2014: Exploration and Field Development in East Siberia},
             url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-84908494593&partnerID=40&md5=98618700045023e9da8e18a18a90aae8},
        keywords = {Characterization; Constrained optimization; Gas industry; Hydrocarbons; Linear regression; Neural networks; Oil field development; Oil shale; Petroleum reservoir engineering; Petroleum reservoirs; Porosity; Public utilities; Well logging, Correlation coefficient; Linear regression models; Multivariable regression; Oil and gas companies; Petrophysical properties; Polynomial neural networks; Probabilistic neural networks; Reservoir characterization, Seismology},
        abstract = {The Main objective of oil industry worldwide is determination of accurate reservoir model. These models make an increased percentage of the world's hydrocarbon reserves 1. A properly constrained reservoir model identifies hydrocarbons and optimizing its production. Accurate reservoir model requires complete information of subsurface properties such as porosity, permeability, etc 9. But the fundamental challenges for geologists and geophysicists to predict these properties are reservoir specificity and heterogeneity which affects reservoir performance and their well productivity. One of the key parameter for identifying hydrocarbon is porosity, but its prediction is difficult because of significant variation over a reservoir volume. The better way to investigating such variation is improvement in spatial distribution of porosity by implementing the integration between the 3-D seismic and well attributes. Moreover, nonlinear multivariable regression technique like Probabilistic Neural Network has been utilizes to correlate statistically the seismic attribute to achieve high correlation coefficients when cross-plotted with reservoir properties. It results in better (r2 = 0.82) correlation coefficient than linear regression model showed (r2= 0.74) 10-11. The challenge to the technique is better seismic-well tie to generate synthetic seismic traces and their correlation between predicted and the true seismic trace. Therefore, we can propose to generate pseudo porosity log from the 3-D seismic volume using polynomial neural network (PNN), helps in better integration between seismic attribute and well logs to improve the reservoir characterization by providing information about petrophysical properties away from well controls. Seismic inversion method is adapted for extracting correlation weights, that implementing on whole seismic volume to produce pseudo log volume. The proposed model tries to achieve high attribute correlation between the predicted and actual porosity logs that are similar to true porosity logs which improve the reservoir characterization that leads in estimating hydrocarbon reserves. This model also assists oil and gas companies to obtain higher drilling success.},
          author = {Kumar, A. and Yusoff, W. I. W. and Asirvadam, V. S. and Dass, S. C.},
            isbn = {9781634391665}
}