Datapages, Inc.Print this page

Seismic Visualization of Thin-Beds Guided by Seismic Geomorphology and Waveform Analysis in Wheeler (Stratal-Slice) Domain

Hongliu Zeng
BEG, The University of Texas, Austin, TX


Seismic visualization of thin-bedded reservoirs is difficult because (1) the top and base of a thin bed is not resolved by seismic data, and (2) the length of seismic response window is significantly larger than the thin-bed thickness. As a result, 3D seismic image of a thin bed is at most a half-true representation of the reservoir. Especially in a realistic stratigraphic model (figure upper) where multiple tightly-spaced thin beds lead to a seismic record of heavy interference (figure middle), to interpret and visualize the individual reservoirs and their spatial relationship is extremely challenging, if not impossible, as illustrated in the figure (lower).

I introduce a seismic sedimentology-based workflow that would shade lights on dealing with this difficult problem for improved resolution of stratigraphic analysis of thin beds. The workflow starts with creating stratal slices for amplitude display of thin beds in Wheeler domain. The thin-bed images can be interpreted in terms of seismic geomorphology for their depositional (facies) meaning (in this case fluvial and distributary channels). However, the superimposed facies patterns of neighboring units on stratal slice prevent us from a clear interpretation of individual systems and their spatial relationships (stacking pattern). A waveform analysis of each facies (depositional element) along relative geologic-time axis can help restore waveform for each of the individual facies. The center of a thin bed can be picked at the maximum amplitude (for 90-degree data) or zero-crossing (zero-phase data). The systematic traveltime differences between individual thin-bed waveforms reveal possible depositional history (stacking pattern).

This is mostly a model-based study; some field data examples will be presented to support the conclusion.

Figure. Upper: a multi-thin-bed impedance model and individual patterns recognized as fluvial (#1 and #2 sandstones) or distributary (#3-6 sandstones) channel systems. Middle: synthetic section made with 25-Hz dominant frequency, 90-degree Ricker wavelet. Lower: amplitude slice picked at #3 sandstone unit.

AAPG Search and Discovery Article #90206 © AAPG Hedberg Conference, Interpretation Visualization in the Petroleum Industry, Houston, Texas, June 1-4, 2014