Characterization and probabilistic estimation of tight carbonate reservoir properties using quantitative geophysical approach: a case study from a mature gas field in the Middle Indus Basin of Pakistan

Abstract

In this study a tight carbonate gas reservoir of early Eocene (S1 formation) is studied for litho-facies estimation and probabilistic estimation of reservoir properties prediction using quantitative geophysical approach from a mature gas field in the Middle Indus Basin, onshore Pakistan. Quantitative seismic reservoir characterization approach relied on well based litho-facies re-classification, Amplitude Variation with Offset (AVO) attributes analysis and Pre-Stack simultaneous inversion attributes constrained with customized well-log and seismic data (gathers) conditioning. Three main litho-facies (hydrocarbon bearing limestone, tight limestone and shale) are classified estimated based on the precise analysis of well data using petrophysical properties. AVO attributes (intercept and gradient) conveniently inspection for amplitude behavior (reflection coefficients) of the possible AVO (class I), fluids and lithology characteristics. Probable litho-facies (tight limestone and shale) are estimated using well based litho-facies classification and inverted seismic attributes (p-impedance and density) from pre-stack simultaneous inversion in a Bayesian framework. Additionally, petrophysical properties (clay volume and porosity) are derived from probabilistic neural network approach using well logs and pre-stack inverted attributes (pimpedance and density) constrained with sample-based seismic attributes (instantaneous, windowed frequency, filters, derivatives, integrated and time).

Exploring the untapped gas potential of Ghazij shale in Pirkoh area, Pakistan: Integrated approach of Attribute analysis and maturity modeling

“…Hence, there might be a possibility that the gas has been generated from shales of the Ghazij Formation at the depth of 4000 m, at the temperature ranged from 120 °C to 150 °C. Such temperature conditions are favorable for the gas generation (Durrani et al, 2020).…”

Bayesian stochastic inversion with petro-elastic relation to quantify thin gas sands of Khadro Formation, Zamzama gas field

“…A Bayesian stochastic inversion is employed to update the stratigraphic grid of the prior models for jointly distributed posterior elastic volumes by incorporating horizontal and vertical variograms of high-frequency well logs with synthetic trends corresponding to the partial seismic angle stacks. Furthermore, using the kernel density technique, the posterior volumes are integrated with modeled elastic logs and well-based lithologies on 2D crossplots to generate the Probability Density Functions (PDFs) and estimate the litho-facies probability cubes in Bayesian inference (Durrani et al, 2020).…”

Gas-bearing sands appraisal for Zamzama Gas field in Pakistan through inverted elastic attributes assisted with PNN approximation of petrophysical properties

“…However, their reliability can be of a certain limit due to variations in the reservoir properties. The advancement in computational technology suggests non-linear relations for solving complex problems even for shale plays without any prior information (Durrani et al, 2020). The Probabilistic Neural Networking (PNN) is a nonlinear, interpolation mathematical procedure, explained in detail by Sinaga et al (2019), which trains the input petrophysical logs with internal (sample-based seismic volume) and external attributes (impedances and Vp/Vs ratio).…”

“…For petrophysical distributions (porosity and clay), Probabilistic Neural Networking (PNN) is used to better manage shales inside sands (Durrani et al, 2020). The main aims include de-risking of the new drilling points and improved reserve scheming using integrated geophysical exploration techniques (Li et al, 2020).…”