A cloud-based Well Log Database Prototype

Abstract

Geoscience data including seismic, well log, sensor and core measurements are fundamental for Petroleum exploration. Due to recent advancements in sensor and computer technologies, the volume of this data is constantly increasing. Having a unified repository of this data of various types, structure and complexity (Big Data) is crucial for maintaining data integrity. This study addresses petroleum exploration data integrity issues. Current trends in data management technologies and current data practices in Petroleum Geoscience are explored and a practical data management solution to facilitate data access, storage and sharing is recommended. A prototype of a well log database was developed to demonstrate the advantage of having a unified repository of downloaded and sanitized data for multiple users. In comparison with current practices, such a database will prevent duplicate downloads from public websites by petro physicists and make data use more efficient within a particular organisation. The prototype was developed using cloud-based technology and the PAWSEY supercomputing facility (a joint venture of CSIRO with Western Australian universities) for storing both the raw (.las and .DLIS files.) and the sanitized well log datasets from Bonaparte Basin. PostgreSQL database was used to store the sanitized well log data, metadata and links to raw data. PostgreSQL architecture was selected for its ability to support advanced data types (arrays, JSON etc.), plug in to languages like Python, and link to PostGIS, a spatial database extender. A web-based graphical user interface was developed to view, upload and download well log data. In addition meaningful metadata standards were established in collaboration with expert petro physicists.