DHI evaluation by combining rock physics simulation and statistical techniques for fluid identification of Cambrian-to-Cretaceous clastic reservoirs in Pakistan |
| |
| Authors: | Nisar Ahmed Perveiz Khalid Hafiz Muhammad Bilal Shafi Patrick Connolly |
| |
| Institution: | 1.Institute of Geology,University of the Punjab,Lahore,Pakistan;2.Patrick Connolly Associates Ltd.,Berkshire,UK |
| |
| Abstract: | The use of seismic direct hydrocarbon indicators is very common in exploration and reservoir development to minimise exploration risk and to optimise the location of production wells. DHIs can be enhanced using AVO methods to calculate seismic attributes that approximate relative elastic properties. In this study, we analyse the sensitivity to pore fluid changes of a range of elastic properties by combining rock physics studies and statistical techniques and determine which provide the best basis for DHIs. Gassmann fluid substitution is applied to the well log data and various elastic properties are evaluated by measuring the degree of separation that they achieve between gas sands and wet sands. The method has been applied successfully to well log data from proven reservoirs in three different siliciclastic environments of Cambrian, Jurassic, and Cretaceous ages. We have quantified the sensitivity of various elastic properties such as acoustic and extended elastic (EEI) impedances, elastic moduli (K sat and K sat–μ), lambda–mu–rho method (λρ and μρ), P-to-S-wave velocity ratio (V P/V S), and Poisson’s ratio (σ) at fully gas/water saturation scenarios. The results are strongly dependent on the local geological settings and our modeling demonstrates that for Cambrian and Cretaceous reservoirs, K sat–μ, EEI, V P/V S, and σ are more sensitive to pore fluids (gas/water). For the Jurassic reservoir, the sensitivity of all elastic and seismic properties to pore fluid reduces due to high overburden pressure and the resultant low porosity. Fluid indicators are evaluated using two metrics: a fluid indicator coefficient based on a Gaussian model and an overlap coefficient which makes no assumptions about a distribution model. This study will provide a potential way to identify gas sand zones in future exploration. |
| |
| Keywords: | |
| 本文献已被 SpringerLink 等数据库收录！ |
|
