Reservoir models are applied to characterize uncertainty in reservoir volumetric and flow rates. These models are constructed under the constraints of data conditioning and available geostatistical methods. There is limited bandwidth for integration of the wealth of information from physics-based, experimental–based and observational-based geological studies. In reservoir analog models, such as stratigraphic forward models, geological concepts can be prioritized over precise data conditioning. These models can be statistically resampled to learn about value of information and relationships between variables of interest.
An example is presented based on a DionisosFlow forward stratigraphic model of a deepwater slope-fan depositional system. Resampling is applied to calculate local thickness, sedimentation rate and hiatal frequencies as they may be related to local reservoir quality and flow performance. From the resampling we have calculated various statistical models that demonstrate: (1) the local variations in the metrics that aid in assigning uncertainty to these measures, (2) the relationships between variables that may be applied to constraint reservoir modeling property distributions, and (3) the value of information from the perspective of how much data is required to characterize these measures. This workflow and example provide an improved method to integrate geological information through reservoir analog models into reservoir models for improved decision making.
AAPG Datapages/Search and Discovery Article #90323 ©2018 AAPG Annual Convention and Exhibition, Salt Lake City, Utah, May 20-23, 2018