Abstract
Accurate assessments of production forecast of field with time is the ultimate goal of any field operator. To achieve this operators employ a multi-disciplinary team approach and try to obtain a unified representation of the field. The data and the resulting model goes from one team to another after getting the best case or satisfactory results and this has been a standard exercise in the industry till now. In traditional approach, the uncertainties at each stage of the modeling process is not carried over to the next stage and as a result, it is lost at each stage of the process. Big loop addresses this gap in the modeling process where seismic to simulation process is dealt with not just one model at a time, but with ensemble of models and with this approach, the uncertainties are not lost in the process but carried over to the next stage and can be quantified in this process from where they occur to where they matter.
The Big loop workflow is automated and can be run many times and updated with additional information obtained like drilling new wells, as the field passes thru' its lifecycle. With the use of stochastic proxy modeling technique, the stochastic uncertainty in the geological model can be captured. The Big loop workflow gives a geologically consistent history matched model without any local modification in the geological model through the use of machine learning approach and proxy model built by the tool. The model can later be extended to prediction times to capture and quantify uncertainty in the prediction phase of the study. The workflow allows the team to spend more time in analyzing the model to build a common understanding rather than wasting time on manual adjustments or spending time on non-influential uncertain parameters which might not have much influence on the parameters under study.
