Optimal Improved Oil Recovery (IOR) depends significantly on the ability to estimate volumes and locations of bypassed oil from available historical data. Assisted simulation history-matching techniques are being used to estimate remaining reserves volumes and locations. This paper presents an approach to history match that more accurately captures model uncertainty. The novelty lies in direct interfacing between the geological modeling software and a forward simulator with the rapid generation of model updates in wave-number domain.

We describe a Bayesian workflow based on two-step Markov chain Monte Carlo (MCMC) inversion. Arguably, the MCMC methods statistically provide the most rigorous way of sampling posterior distribution, but when deployed in direct simulation, suffer from high computational cost. We outline an approach where the proxy model is guided by streamline-based sensitivities, dispensing with the need to run forward simulation for every model realization. We generate an ensemble of sufficiently diverse static model realizations at the high-resolution geological scale that generates more accurate results by obeying known geostatistics (variograms) and well constraints. An efficient model parameterization and updating is described for rapid characterization of the main features of geologic uncertainty space: structural framework, stratigraphic layering, facies distribution, and petrophysical properties.

We validate the workflow on a case study combining geological model with ~900k cells, four different depositional environments and 30 wells with 10-year water-flood history. The method can be used for successfully identifying the highest potential to capture bypassed oil and for implementing IOR. A history-match method is presented that has potential to lead to better IOR decisions and results through more accurate simulation models and inherent quantification of uncertainty.

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