Prior falsification of conceptual models using well data and seismic data

Céline Scheidt Suihong Song Tapan Mukerji
Realistically assessing reservoir uncertainty is an important challenge in reservoir forecasting. At the appraisal phase, current practice consists of building a set of models from seismic and well data and then using these models to assess uncertainty on some prediction variables and create models that match the available data (inverse problem). Because of high uncertainty, a variety of conceptual models should be used initially. The focus then lies on rejecting scenarios that are incompatible with data using some measure based on the fit of the predictions to the actual observed data. The proposed measure, based on the wavelet transform decomposition, analyzes differences in patterns found in the seismic data and aims at comparing seismic images as a whole, avoiding any local trace-by-trace comparison. Based on this global measure of similarity, our procedure identifies the geologic interpretations that are unlikely given the observed seismic data. Different measures are investigated to ensure robustness of the procedure. The scenarios that are not falsified constitute the solution of the inverse problem.