Structural uncertainty is one of the most significant uncertainty affecting hydrocarbon volume in place and the dynamic behavior as well. A reduction of reservoir geometry uncertainty is desirable during the history matching phase; however, it is in general hard to achieve with standard techniques. On the other hand, ensemble history matching is becoming the trend of the history matching methodologies, especially using the Ensemble Smoother with Multiple Data Assimilation techniques (ES-MDA) since it allows a proper evaluation of post-match uncertainty.
The main objective of this work is to assess the possibility to combine a suitable parameterization of the structural uncertainty with the ensemble history matching process. The result is a practical workflow, which evaluates the impact of structural uncertainty on the post-match static hydrocarbon volume in-place and reservoir model-based production forecasts.
Structural uncertainty has been parameterized using a surface to quantify possible reservoir shape variations and then applying an elastic gridding concept. The work has been performed inside a standard reservoir geo-modelling tool in order to create full 3D structural modifications representing alternative reservoir geometries. The structural uncertainty has been then integrated in an ensemble approach, developing a methodological workflow, which integrates in the data assimilation step the geo-modelling tool with a commercial implementation of the ES-MDA.
An application of the workflow shows that, including structural uncertainty parameterization, quality of history matching increases and post-match uncertainty decreases.
In addition, structural uncertainty post-match modifications could be linked to other history matching parameters with significant volumetric impact. Indeed, different matched alternatives can be found by different combinations of modified geometries and other relevant volumetric parameters, like porosity or Net-To-Gross (NTG), giving a better quantification of post-match forecasts uncertainty.
History matching uncertainty parameterization is usually based on scalar values representing, for instance, transmissibility of faults and, in the ensemble case, on the 3D grid properties like the porosity, permeability and the distribution of other other properties. However, the structural uncertainty has also an impact of both static volumes and ultimately productions forecasts; therefore, both the quantification of the structural uncertainty and its calibration through History matching process are crucial (Seiler, Rivenaes, Aenonsen, & Evensen, 2009).
