Uncertainty modeling with the study strip

With the study strip, you can set up studies, which are an assembly of settings that steer an uncertainty run. This can either be a volumetric uncertainty calculation, or a dynamic simulation incorporating uncertainties.

Input to a study is always a concept, which is created with the CONCEPT strip and contains the sequence of domain models that are called upon when running the study. These domain models contain parametric uncertainties in the form of probability density functions (PDFs) or seeds, specified in the various domain model workflows (e.g., porosity model, fluid model or structural model).

For volumetric uncertainty calculations, you specify whether the parametric uncertainties are independent or correlated. In the Volumetric Study workflow, you then construct realizations by sampling these PDFs, using one of the available sampling techniques. Each uncertainty parameter receives one sample per realization, after which the computational model can be run with a single value for each parameter distribution. In this way the probability distributions are propagated to your calculation results.

For dynamic simulation that incorporates uncertainty, setting up and running the study is slightly different from a volumetric uncertainty calculation. The input consists of a concept, but instead of JewelSuite performing the sampling of the uncertainty parameters and running the study, this process is handled by CMG's optimization and analysis tool CMOST. The simulations are executed by the CMG simulators, after which the results of the uncertainty run are sent back to CMOST for analysis.