The scope of this publication is to capture the main learnings from the application of ensemble-based modelling of three giant geologically complex carbonate reservoirs onshore and offshore Abu Dhabi, simultaneously considering static and dynamic uncertainties. The ability to consider these uncertainties in prediction studies is highlighted, leading to optimal economic decisions to be taken in the subsequent stages of development of these critical assets.
For each oilfield, an integrated static-to-simulation modelling workflow was built in collaboration with the relevant asset teams, capturing their knowledge and expertise in the generation of an ensemble of cases, equiprobable and plausible from the point-of-view of geology and dynamic characteristics. Each of them has specific geological and hydrodynamic challenges to be taken into account, from the spatial distribution of the static rock types and of their heterogeneous petrophysical properties, the impact on the flow of high permeability streaks and stylolites, to the behavior of the aquifers. Each ensemble of cases is subsequently utilized to assimilate the production data using an iterative ensemble Kalman method yielding ensembles with the ability to reproduce the observed reservoir dynamics. These calibrated ensembles can subsequently be used for predictions and economic evaluations considering all remaining static and dynamic uncertainties. After data assimilation, the ensembles showed reasonable match to field and well historical data for the three different studies. There were significant learnings in the static and dynamic updates and uncertainty reduction that occurred during the data assimilation. They provided statistical insights with respect to the reservoir characterization, such as increased high permeability streaks probability in specific zones or reduced uncertainty surrounding the porosity/permeability transforms for each rock type, and fluid dynamic, such as fault behavior. These learnings will benefit the team to further their understanding and improve future modelling activities. Multiple development scenarios were considered for each asset and the simulated ensemble results were brought for economic evaluation under static and dynamic uncertainties. This provided representative estimates of the net present value of each scenario and eventually, a complete understanding of the potential outcome, allowing for informed decisions. Finally, another important benefit of working with calibrated ensembles was shown in its ability to identify the most likely bypassed area from a probabilistic standpoint, allowing to take confident decisions for new target identifications to increase the ultimate field recovery.
While addressing the future challenges of major carbonate oilfield developments and to ensure an optimal decision-making process, the asset team has to consider the complexity of the underlying geological environment, the dynamics of the fluid in the reservoir and their associated uncertainties. An integrated ensemble-based approach from static to simulation with fast data assimilation and economic evaluation of possible scenarios proved to be key to reach all of these objectives.
