The increased amount of available measurements when drilling with wired pipe opens new possibilities for well placement optimisation. This requires effective interpretation, integration and utilisation of the new information within the timeframe set by the ongoing drilling operation. Earth models carry knowledge about geological structures and physical properties, and form part of the basis for geosteering applications. However, current earth modelling tools have limited capabilities for effective integration of newly acquired information. Model modifications are complex and labour intensive, and the time needed exceeds the time available during drilling. This is a large drawback for decision-making processes that require the most current and precise information.
The complexity of the earth model modification process in existing methodologies is partially a result of storing physical parameters in grids. The grids are controlled by a structural model composed of fault and horizon geometries obtained from geological and seismic interpretation.
We have developed a gridless strategy based on separation of structural information from physical parameters. The separation of the geometric and the parameter models enables modifications to be handled individually and more efficiently. It increases control with model resolution which allows improved management of large data sets, access to earth model information at any scale and efficient sharing over computer networks.
The main principles of our approach are demonstrated through geologically realistic test cases that imitate real drilling scenarios in a reservoir with normal faulting. The combination of new technologies may form a basis for closed loop control/automation of sub processes, required for future automation of the whole drilling process.