--> ABSTRACT: Lithology and Fluid Prediction: Well Ties and the “Rock-Physics Bottleneck”, by Cooper, Richard, Matthew Carr, Naum Derzhi, M. Turhan Taner, Richard Uden, Jack Petrovich Dvorkin, Joel Walls, Gary Mavko; #90026 (2004)

Datapages, Inc.Print this page

Cooper, Richard1, Matthew Carr1, Naum Derzhi1, M. Turhan Taner1, Richard Uden1, Jack Petrovich Dvorkin1, Joel Walls1, Gary Mavko2
(1) Rock Solid Images, Houston, TX
(2) Stanford University, Stanford, CA

ABSTRACT: Lithology and Fluid Prediction: Well Ties and the “Rock-Physics Bottleneck”

Our industry is making increasing use of 3D seismic data for lithology, fluid, and porosity prediction. The primary source of seismic calibration data is well logs. Hence we must rely on the synthetic seismogram as the primary interface between seismic data, logs, and reservoir properties. Quality of tie between the synthetic seismogram and the seismic data is a major factor in our ability to robustly calibrate seismic data to rock and fluid properties.

For geophysical modeling, we are interested primarily in accurate Vp, Vs, and density. Indeed, seismic inversion can yield only these acoustic attributes (and possibly attenuation). This critical dependency upon a small number of measurements is referred to as the “rock physics bottleneck” (Mavko). Most well log analysis is focused on formation evaluation, but for reservoir geophysics, many more issues must be considered.

Some of those issues are:
Inadequate vertical coverage of log data
Inadequate or missing sonic, density, and dipole-sonic logs.
Drilling mud invasion
Synthetics at all offsets or zero-offset only?
Ray-tracing or full-elastic modeling (speed vs. fidelity tradeoff)?
All arrivals or primaries only?
Attenuation and dispersion.
Positioning errors
Sampling, upscaling, and resolution

This paper shows several examples that illustrate the key issues to consider when constructing synthetic seismic models to aid in seismic reservoir characterization. It will be necessary for our industry to consistently address each of these factors to ensure we deliver robust and reliable reservoir models derived from geophysical data.

 

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