Case Study: Shell achieves rapid, highly accurate subsurface reservoir predictions with MathWorks' MATLAB

Shell Geologists Develop and Deploy Software for Predicting Subsurface Geologic Features

Shell geologists at Shell faced the challenge of predicting subsurface geologic features that seismic images could not fully resolve, a problem that can lead to costly unsuccessful wells. They wanted to develop and control their own predictive analytics in MATLAB and deploy the results across the company, so they partnered with MathWorks and used MATLAB and related toolboxes to build their stratigraphic characterization and analog-matching algorithms.

MathWorks consultants helped refine the MATLAB code, add a MongoDB connector and web mapping, and package and deploy the solution with MATLAB Compiler SDK and MATLAB Production Server. The resulting application delivers answers in minutes instead of months, produced a drilling prognosis accurate to within 12% (compared with industry errors often around 1000%), and lets Shell instantly roll out algorithm and database updates to users worldwide.

Shell

Nick Howes

Geologist