A Computer Application for Automatic Interpretation of Depositional Process from Visual Descriptions of Sedimentary Facies

Carbonera, Joel L.; Rosa, Douglas E.; Abel, Mara; Scherer, Claiton M.

The petroleum exploration activities usually require the performing of several data intensive tasks. For example, the description of outcrops and cores, and the interpretation of depositional process, are key tasks for the reconstruction of the stratigraphic history of a given area and, therefore, they are considered as important tasks in the exploration phase. These tasks are resource-consuming and, due to this, there is an increasing interest in the development of intelligent software that automate them. In this context, we developed an intelligent system for automatic interpretation of depositional processes. Our system takes a visual description of sedimentary facies as input and interprets the depositional process that have been responsible by its generation. A sedimentary facies is a region in a core or outcrop, which is visually distinguishable of adjacent regions. Depositional processes are events that involve a complex interaction of natural forces and sediments, and which are responsible for the formation of sedimentary rocks. In order to ensure the reliability of our software, we consider the necessity of qualifying the acquisition of the visual descriptions of facies. This was ensured through the development of an expressive formal vocabulary that represents the concepts applied by geologists when they talk about the visual characterization of sedimentary facies. Once the user supply the system with a facies description, using this formal vocabulary, an interpretation engine perform the inference of the corresponding depositional process. To allow this inference, we feed the interpretation engine with the expert knowledge which relates visual features of sedimentary facies (evidences) to depositional processes (interpretations). At the current stage, the software was tested on tasks of interpretation of tractive currents. The performance of the software was evaluated with three cases taken from the literature (peer reviewed papers) of Sedimentary Stratigraphy: northwestern part of Kyongsang basin (South Korea), Warchha Sandstone of the Salt Range (Pakistan) and Barreiras formation (Brazil). In these cases, our system demonstrated a total of 55,17% of true positives, 44,83% of true negatives, without either false positives or false negatives. In the future, our system will be integrated with Strataledge® system for sedimentary facies description in cores and outcrops.

AAPG Search and Discovery Article #90163©2013AAPG 2013 Annual Convention and Exhibition, Pittsburgh, Pennsylvania, May 19-22, 2013