Prediction of Petrophysical Properties Through Comparative Post-Stack Inversion Techniques Using Advance Neural Networking

DOI : 10.48129/kjs.18279

Authors

  • Pal Washa Shahzad Rathore Department of Earth Sciences, Quaid-i-Azam University, Islamabad, Pakistan
  • Matloob Hussain Department of Earth Sciences, Quaid-i-Azam University, Islamabad, Pakistan
  • Muhammad Bilal Malik Department of Earth Sciences, Quaid-i-Azam University, Islamabad, Pakistan
  • Sher Afgan Institute of Geology, University of the Punjab, Lahore 54000, Pakistan

DOI:

https://doi.org/10.48129/kjs.18279

Abstract

The sophisticated seismic inversion methodology can develop the relationship between the interpreted seismic data and the elastic properties of the reservoir. The comparative analysis of three seismic inversion techniques, Band limited inversion, Model Based inversion and stochastic inversion, is used to assess each technique's efficiency. The comparison also helps better understand reservoir petrophysical properties (Vclay and Effective porosity). The reservoir in the study area is the C-interval sands of the Lower Goru Formation. Data used in the present research work include 3D seismic cube and bore-hole logs data of six wells drilled in the study area. For the calculation of Petrophysical properties (Vclay and Effective porosity), inverted attributes are used as an input in Probability Neural Networking along with Post-stack Time Migration. The is a useful technique for predicting desired petrophysics properties in sands encased within shales. The comparison of results from all three techniques after PNN is analyzed to achieve the study's goal. The results of Probability Neural Networking using stochastic inversion attribute match the blind well and resolute delicately the reservoir potential along with populating the depositional environment of the three segregated sand bodies. In the present study, the model is trained to predict well location using the petrophysical well logs data, Inverted Impedance, and seismic data. After the training of the model, it is used to predict the entire seismic cube.

Published

04-08-2022

Issue

Section

Earth & Environment