Advanced Methods to Build Complex Earth Models for Seismic Modeling

Abstract

Abstract

Introduction

Finite-difference modelling has recently become a powerful instrument to accurately simulate seismic wave propagation. The algorithm can handle anisotropic, viscous, and structurally complex models, and, as such, has triggered an interest in the development of new techniques to build earth models for modeling purposes. Herein, we present some of these new techniques.

Description

Model building is generally performed in two steps: generation of a structural framework followed by population of properties into the framework.

Surfaces derived from interpreted horizons are edited and assembled together to produce the structural framework. It is also possible to introduce interpreted faults into the framework. The process can be challenging if a large number of faults is introduced. This technique was recently adopted to create a complex model with more than ten faults representing an overthrust system (Figure 1a). The model was built as part of a modeling study to evaluate the design of an acquisition geometry in an area of the Gulf of Mexico.

The propagation velocity and the rock density are the main properties required to generate finite-difference modeling synthetic data. The velocity and density values can be extracted from available sonic and density logs, respectively, and propagated through the framework (Figures 1b and 1c). Additional properties, such as the epsilon anisotropy (Figure 1d) and the attenuation can also be estimated and incorporated into the model.

If well logs are not available, it is also possible to extract seismic amplitudes from the migrated seismic volume (Figure 2d) and derive an estimation of the impedance and density contrasts. This technique was recently implemented to populate the density property into a complex structural model (Figures 2a and 2c) in the Alaminos Canyon area of the Gulf of Mexico.

A section is selected to generate and migrate 2D finite-difference modelling synthetics gathers. The resulting 2D image (Figures 2c) is compared to the real seismic image (Figure 2d) as part of the model validation process. A few iterations are required to adjust and finalize the model.

AAPG Datapages/Search and Discovery Article #90260 © 2016 AAPG/SEG International Conference & Exhibition, Cancun, Mexico, September 6-9, 2016