--> Abstract: Using Simple Validation Techniques to Minimise Uncertainty When Assessing Multiple Structural Interpretations; #90063 (2007)

Datapages, Inc.Print this page

Using Simple Validation Techniques to Minimise Uncertainty When Assessing Multiple Structural Interpretations

 

Bond, Clare E.1, Alan Gibbs2, Zoe K. Shipton1, Serena Jones3 (1) Glasgow University, Glasgow, United Kingdom (2) Midland Valley Exploration Ltd, Glasgow, United Kingdom (3) Midland Valley Exploration Limited, Glasgow, United Kingdom

 

Unique structural solutions that honour sub-surface data are rare, particularly for areas of poor data quality or complex structural geometries. Assessing and capturing the full range of possible structural solutions to a 2D or 3D seismic dataset is challenging. Perhaps more critical is that, within the current oil and gas workflow, we reduce all of these possible interpretations to a single model for development at the reservoir engineering stage. These single model solutions to complex and ‘fuzzy' datasets are used to create highly refined reservoir models that have associated uncertainty models for petrophysical and other reservoir attributes. These uncertainty calculations are produced with little or no consideration to the original validity and uncertainty in the structural model.

 

We have asked over 400 geoscientists to interpret a 2D seismic section, and have collated the results to capture the range of concept models and structural interpretations. The results show that a wide range of structural and concept models can be fitted to the data, but many are geometrically and kinematically invalid. Analysis of the interpretation process shows that the most successful geoscientists demonstrate an ability to think in the 4th dimension (time) and to query their own interpretation for evolutionary validity. These abilities have a dramatic affect on model outcome and on the geoscientist's confidence in their model (an uncertainty indicator). Our experiment shows that the use of simple evolutionary validation techniques at the initial interpretation stage reduces the number of erroneous concept models for the data and hence decreases structural uncertainty.

 

AAPG Search and Discover Article #90063©2007 AAPG Annual Convention, Long Beach, California