Subsurface Data Solutions

Our research and development activities in the past few years has been focused on the use of well-established open source technologies for holding massive amounts of data and being able to access them dynamically with infinite scalability. Two questions that have been uncovered in the industry have driven us to focus on this problem. Both of these are part of a larger challenge - and need - for a unified data store at the back end for all exploration geophysics, drilling, and production data. Filling these two gaps will go a long way to addressing the bigger challenge.

The first one is about aggregating the surface and subsurface log data from all wells into a harmonized, normalized data repository so that analytics can be deployed to look for and catalog patterns of interest so that they may be found on demand as new heuristic models are tested and early warning systems are developed.

The second challenge is about creating a single addressable store for seismic data and being able to generate clipped SEGY files on demand for specific horizons or small shape files so that geophysicists can rapidly access and load the data of interest.

Both of these are addressed with the new data stores we have developed for the industry using the Apache™ Hadoop™ framework of open source software known for its reliability and scalability.

Drilling Data Server & Repository (DDSR™) is the unified data store foundation for the analytics used to assess the patterns of interest in the logs as well as for text parsing, natural language processing (for reports, drilling plans, and reports) so that the entire well can be described by its attribute metadata in a structured, referenceable manner. The extracted metadata is also housed in the unified data store and made available for configuring future models for predictive safety or performance optimization.

Seismic Data Server & Repository (SDSR™) supports the needs of mid-sized and large E&P companies (IOCs), NOCs, and National Data Repositories. SDSR stores post-stack data in SEGY format and provisions this data to the interpretation tools running on workstations – anytime, anywhere. Pre-stack and post-stack data is also stored, leveraging the scalability of the open source Hadoop environment.

Geotechnical Marker Detection & Analysis
PointCross provides facilities for extracting potentially interesting patterns in the G&G data using analytics and statistical packages. We also provide the means to catalog and index these patterns used by detection algorithms to serve specific search criteria that users develop. Pattern extraction is done for patterns that are potentially usable in creating likely scenarios of interest and where these scenarios have a high probability of predicting an event or an occurrence of interest.

The reason for extracting patterns and the way they are used on historical data to learn and develop models which can be then used for decision support in new wells being drilled follows a sequence of Observe-> Analyze -> Model -> Test. We are working closely with industry experts and customers to build an extensible library of templates and reusable geotechnical-markers and search panels.

Geotechnical-marker Library –This library allows users to build, classify, and catalog their own geotechnical-marker and test them on the data. They may be published to the larger user group and they are combined with standard panels provided by PointCross. Each of the markers includes several typically measured parameters and allowable levels or patterns. They may be time series or combinations of parameters.