Abstract
Reservoir structural modelling is one of the fundamental steps in a reservoir study workflow. The impact of the structural uncertainties on the dynamic response of the reservoir is well known and not negligible, but often the reservoir shape is considered as fixed due to the complexity to manage alternative geological structures in multi-realisations simulation loop. Nevertheless, both Risk Analysis (RA) and History Matching (HM) workflows strongly require a practical and time-effective methodology for structure management with an efficient uncertain geometry parameterization. In this work, an innovative methodology for structural uncertainty handling is presented. The methodology is based on the combination of Principal Component Analysis (PCA) and Elastic Gridding. In particular, the PCA-based parameterization is able to efficiently handle the geophysical uncertainty model, consistent with the geostatistical characterization as well. Such methodology has been structured in an internally developed tool. This tool is specifically designed for a direct handling of corner point geometry grids and allows changes of surfaces, shape and size of internal reservoir layers, fault throw and fault position and even new fault placement, honouring geological constraints. One of the key points of the proposed methodology is the integration of a geologically-oriented parameterization and a statistical parameters reduction technique (the above mentioned PCA) in a workflow which includes commercial HM/RA tool and a dynamic simulator. The result is an efficient structural uncertainty management framework suitable for Risk Analysis and History Matching studies. Among the field applications performed so far, two cases have been chosen aiming at showing the potentialities of the proposed approach. The first example is a history matching exercise on an undersatured oil reservoir. A comparison between the traditional and the "structural", even if simplified, HM is herein provided, showing the improvement due to a better geologically-oriented uncertainty model. The second example is a risk analysis application on an oil field, with a strong uncertainty of the oil in place due to lack of accurate knowledge of the reservoir flanks shape. The application highlights the advantages deriving from the geophysical PCA-based workflow.
