Machine-assisted 4D seismic interpretation

We investigate methods and workflow to automate parts of the 4D interpretation workflow. We detect reservoir changes and use machine learning to predict cause of the 4D seismic effects. The benefits are expected to come from reduced overall time needed for 4D seismic interpretation, and faster integration of insights into well target identification, well planning and reservoir model updates.We demonstrate the data-driven method and workflow with examples from the Snorre field, explain some of the challenges met, and end with indicating some of the possibilities for value creation that open as a result of the method.