Comparing different approaches of time-lapse seismic inversion

Daiane R Rosa,Juliana M C Santos,Dario Grana,Denis J Schiozer,Rafael M Souza,Alessandra Davolio,Yanghua Wang

doi:10.1093/jge/gxaa053

Abstract

Abstract Time-lapse (4D) seismic inversion aims to predict changes in elastic rock properties, such as acoustic impedance, from measured seismic amplitude variations due to hydrocarbon production. Possible approaches for 4D seismic inversion include two classes of method: sequential independent 3D inversions and joint inversion of 4D seismic differences. We compare the standard deterministic methods, such as coloured and model-based inversions, and the probabilistic inversion techniques based on a Bayesian approach. The goal is to compare the sequential independent 3D seismic inversions and the joint 4D inversion using the same type of algorithm (Bayesian method) and to benchmark the results to commonly applied algorithms in time-lapse studies. The model property of interest is the ratio of the acoustic impedances, estimated for the monitor, and base surveys at each location in the model. We apply the methods to a synthetic dataset generated based on the Namorado field (offshore southeast Brazil). Using this controlled dataset, we can evaluate properly the results as the true solution is known. The results show that the Bayesian 4D joint inversion, based on the amplitude difference between seismic surveys, provides more accurate results than sequential independent 3D inversion approaches, and these results are consistent with deterministic methods. The Bayesian 4D joint inversion is relatively easy to apply and provides a confidence interval of the predictions.

Highlights

Time-lapse (4D) seismic data are a valuable source of information for monitoring and managing the production of hydrocarbon reservoirs
The processing and interpretation of 4D data include the calculation and evaluation of the changes in seismic attributes resulting from changes in reservoir pressure, fluid saturation and temperature caused by reservoir production ( Johnston 2013)
The limitations of the deterministic methods are highlighted in Francis (2005), in the comparison with realisations of a probabilistic method, showing a significant underprediction of net sands in the deterministic inversion due to the smoother estimated impedances, whereas the probabilistic inversion better reproduces the impedance distribution

Summary

Introduction

Time-lapse (4D) seismic data are a valuable source of information for monitoring and managing the production of hydrocarbon reservoirs. Once 4D seismic data are inverted to estimate elastic properties, the results obtained are assimilated in the dynamic fluid flow simulation workflow. Deterministic methods generally provide a single solution, constrained by a prior model or well data, and include, among others, coloured and model-based inversions (Simm et al 2014). A commonly adopted probabilistic method in reservoir characterisation is the Bayesian inversion, which allows evaluating the impedance of the most likely model and its uncertainty (Buland & Omre 2003) by combining a priori information on model parameters with the measured data in a probabilistic formulation. The limitations of the deterministic methods are highlighted in Francis (2005), in the comparison with realisations of a probabilistic method, showing a significant underprediction of net sands in the deterministic inversion due to the smoother estimated impedances, whereas the probabilistic inversion better reproduces the impedance distribution

Objectives

Results

Conclusion