--> Process-Like Modeling in Turbiditic Channel Environments Constrained by Well Data, 3-D and 4-D Seismic Attributes Application to Offshore West Africa Data

2019 AAPG Annual Convention and Exhibition:

Datapages, Inc.Print this page

Process-Like Modeling in Turbiditic Channel Environments Constrained by Well Data, 3-D and 4-D Seismic Attributes Application to Offshore West Africa Data

Abstract

We used a process-like modeling tool to model sedimentary objects in a turbiditic channel environment. Our objective was to develop a workflow integrating in the tool 4D seismic anomalies as constraints in addition to 3D seismic attribute and well constraints. Tests on West Africa data yielded very encouraging results.

The method we used is an in-house process-like technique, U-like, that models at reservoir scale sedimentary bodies characterized by sediment transport using morphological body characteristics. We established a workflow to model channels and associated levees while honoring geological and geophysical constraints, in particular 4D anomalies.

A deep-offshore West Africa dataset was tested. Its reservoirs are made of laterally and vertically stacked channels with well developed erosive constructive channel fairways with limited levee extensions. A reservoir grid was available with information to constrain the channel and levee modeling per reservoir interval. The seismic-interpreted environments of deposit limits provided the fairway regions where channels were expected. Out of these regions, no channel were simulated. Interpreted elementary channels provided ranges of channel geometry to model (wavelength, amplitude, thickness, concavity). Well facies logs provided the sedimentary objects to honor along boreholes. The lithological classification of inverted 3D seismic data and the geological targets provided the 3D object proportion targets. 4D seismic anomalies provided geobodies where channels were expected.

The workflow developed and tested demonstrated U-like strengths vs. other modeling methods. Results were geologically consistent, object geometries were respected. Vertical, deviated and horizontal well data and 3D object proportions were very well honored (90% on average). The option available to use 4D geobodies directly as a constraint could not be fully tested given the high number of geobodies although very encouraging results were obtained. Instead we integrated the 4D geobodies in the 3D channel proportions by revising proportions at geobodies’ locations. The modeled channels still honored the target proportions. 4D anomalies’ respect is improving the results. The way forward should first consider the 4D signal in modeled levees to address it by facies modelling when expected, then properly automatize the 4D geobodies integration in U-like for direct channel constraints.