Assessing and improving subsurface models is an integral part of any field development project. Scanning through a multitude of data and models to look for opportunities to improve is a challenging task that can be very complex and time consuming if done manually. On the other hand, successful automation of the process typically requires field-specific customization. To address this challenge, a cloud-based innovation toolkit has been developed.

The toolkit is a set of Python modules that automate the key steps in comparing reservoir model predictions to various types of observation data, including production data and time-lapse seismic data. The toolkit enables users to create customized images that highlight important aspects of data and models, helping the team to decide how to proceed. Images that drive decisions are referred to as actionable insights. The toolkit is also used to create customized mismatch metrics, combining, for instance, production data mismatch and time-lapse seismic mismatch per region, per time interval, or overall. Ranking a set of reservoir models according to selected mismatch metrics is automated. For improved efficiency, the toolkit is built on top of a data management system for simulation models, which makes access to the models and results stored in the cloud quick and seamless.

The innovation toolkit was used in a case study in the Grane Field to assess time-lapse seismic data, production data, and two ensembles of 200 and 100 reservoir models, respectively. An important element of the project was the customized co-visualization of gas migration observed from time-lapse seismic data and production data, vs. simulated gas thickness and gas production from the reservoir model ensembles. These images allowed the team to quickly identify a region where changes to the top reservoir structure could reduce the mismatch between model predictions and production and time-lapse seismic observations.

As a final step of this study, an updated ensemble was generated that allowed the team to test and confirm several hypotheses about structure, contacts, and in-place volumes of the selected region of the Grane Field. The short time spent for this study vs. the amount of data processed highlights an important advantage when it comes to the need for continuous well maturation in the Grane Field.

Cloud access to data, models, and new digital tools enable new opportunities for automation and improved efficiency. The novelty of the presented work is the combination of automation and customization, enabling domain knowledge to drive automated workflows.

You can access this article if you purchase or spend a download.