Imhof, Matthias G.¹ 
(1) Virginia Tech, Blacksburg, VA

ABSTRACT: A First Step Towards Quantitative Seismostratigraphic Inversion

The current practice of building reservoir models often patches together separate models for the different reservoir data which can yield inconsistent results without truly bridging scales, resolutions, and coverages. 
Hydrocarbon reservoirs are formed and modified by physical processes which can be formularized with process models. Given data, these models can be inverted to determine the parameters necessary to obtain a reservoir model consistent with these data. This inversion is often called quantitative stratigraphic inversion. Scales, resolutions, and coverages are interpolated by the process models in a geologically sound manner. 
Currently, quantitative stratigraphic inversion is mostly based on interpreted wireline data. Despite excellent coverage and abundance, seismic data have not been used, yet. We are developing an inversion scheme based not only on wireline data, but also on seismic amplitudes and seismic estimates of petrophysical quantities such as porosity or the sand/shale ratio. Initial experiments with seismostratigraphic inversion based on simple models and data are very encouraging, albeit computationally still expensive. We believe, however, that the high computational costs are a minor problem solved by increasing computer speed and grid computing technology.

AAPG Search and Discovery Article #90026©2004 AAPG Annual Meeting, Dallas, Texas, April 18-21, 2004.