Prospect De-Risking Using Seismic Forward Modelling of Frequency RGB Blends

Abstract

Frequency decomposition and RGB blending is a very effective way to highlight geological features in the seismic data. It is a technique that is often used to identify depositional systems and to infer changes in bed thickness. But one of the enduring questions with the interpretation of RGB blend is understanding exactly what the colors mean in terms of the underlying rock properties. In this study we used seismic forward modelling to validate our interpretation of an anomalous high amplitude feature seen within seismic data from the Taranaki basin offshore New Zealand. The feature displayed characteristics and geometries that were indicative of a crevasse splay but we wanted to test this hypotheses using forward modelling. Different models were created with increasing levels of detail and the corresponding reflectivity data and frequency RGB blends were correlated against the original data. This enabled us to understand which elements of the RGB response were controlled by the geometry of the feature and which ones were controlled by rock properties. Varying the rock properties within the model in a manner consistent with crevasse splay deposition (ie. fining up sequences and small scale clinoforms) resulted in a synthetic reflectivity response and frequency RGB blend which closely matched our original data. This gave us greater confidence that the anomaly seen within the data was indeed a crevasse splay and enabled us to understand the distribution of higher quality sands within the anomaly.

AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019