Ensemble-based decision-support workflow for operational geosteering

Abstract

Summary We present a systematic geosteering workflow for decision support under uncertainty. The uncertainty is captured in an earth model represented as an ensemble of realizations of the geology. The realizations are updated each time new real-time measurements become available. Thereafter, the up-to-date ensemble of realizations is used as input for decision support optimization. The optimization algorithm evaluates the well path for the complete drilling operation under uncertainty and proposes a decision that maximizes the expected well value, calculated based on selected objectives. The workflow combines the near-bit information produced from look around with the pre-drill knowledge in realtime to suggest a decision that results in statistically optimal steering. We illustrate this by applying the workflow to two cases; the same initial geomodel is updated in accordance with two different synthetic true models. In both scenarios the workflow successfully augments look-around information with pre-drill knowledge to produce good steering decisions under uncertainty.