Rock physics based non-linear AVO waveform inversion of North Sea greensand

Asger S. Johansen*, Thomas M. Hansen, Ida L. Fabricius

*Corresponding author for this work

Research output: Contribution to book/anthology/report/proceedingArticle in proceedingsResearchpeer-review

Abstract

We present a method for inverting seismic pre-stack angle gather data to obtain a probabilistic estimate of rock physical properties of a hydrocarbon reservoir. The method is based on knowledge about a reservoirs lithology as well as an identifiable reservoir AVO (amplitude variation with offset) response. We make use of a Bayesian formulation of the inverse problem that allows a complex non-linear forward model and describes the properties of the reservoir as an a posteriori probability distribution. We use the extended Metropolis algorithm to sample the a posteriori probability density. The inverse method is studied by inverting seismic data of a greensand hydrocarbon reservoir in the North Sea. The forward model is nonlinear, comprised of a non-linear greensand rock physics model and a non-linear AVO waveform model. The rock physics model is based on a Hertz-Mindlin grain contact model combined with the Gassmann equation for fluid substitution. The inverse problem is parametrized as a 3-layer model, where the intermediate layer is known to be greensand. In a synthetic study we find that the AVO response of greensand is sensitive to the oil saturation. This facilitates the possibility of inverting this property. The physical properties inverted are the oil saturation, glauconite fraction and porosity of greensand. The performance of the inversion is found to be sensitive to the degree of prior information provided and conditional on the effective pressure, the glauconite microporosity and coordination number being defined due to the complexity of the rock physics model. The oil saturation, glauconite fraction and porosity of greensand are successfully inverted from seismic data of the Nini-4 well location, in case the elastic properties of a single layer of the three-layer model are well known a priori. Inversion results are compared to well log and published laboratory core plug data from the field.

Original languageEnglish
Title of host publicationSEG Technical Program Expanded Abstracts 2013
Number of pages5
PublisherSociety of Exploration Geophysicists
Publication date1 Jan 2013
Pages3139-3143
DOIs
Publication statusPublished - 1 Jan 2013
Externally publishedYes
EventSEG Houston 2013 Annual Meeting, SEG 2013 - Houston, United States
Duration: 22 Sept 201127 Sept 2011

Conference

ConferenceSEG Houston 2013 Annual Meeting, SEG 2013
Country/TerritoryUnited States
CityHouston
Period22/09/201127/09/2011
SeriesSEG Technical Program Expanded Abstracts
Volume32
ISSN1052-3812

Fingerprint

Dive into the research topics of 'Rock physics based non-linear AVO waveform inversion of North Sea greensand'. Together they form a unique fingerprint.

Cite this