Abstract

Predicting geology! Whether this prediction concerns reservoir quality in the context of nearby exploration, the level of reservoir heterogeneity in the context of a -development project, or the organization of facies, petrophysical properties, and fracture networks in the context of the installation of development wells, its reliability is a key factor in the economic success of projects. This is obviously true for the hydrocarbon industry, whatever the level of maturity of the fields, but it is also the case for any activity related to the subsurface, such as geothermics or C02 storage. 

For decades, the massive use of geostatistical techniques to populate reservoir models has obscured this need for a predictive capability of the models. While geostatistical models are characterized by their very low predictive capacity, and by the impossibility of testing and validating the concepts to be reproduced, not everyone is aware of the significant limitations and shortcomings of such geostatistics. On the contrary, the advantages associated with its easy use in operational workflows are well known to geologists and reservoir engineers who would not consider not finding such advantages in any alternative technique.

The time has come for techniques capable of actually predicting geological properties to take their appropriate part in the reservoir modelling process. New software applications are arriving, based on geological concepts that feed functionalities combining reduced complexity process-based modeling algorithms with rule-based techniques. These forward geological simulations aim to predict reservoir properties resulting from diverse and successive geological phenomena well beyond the wells and explored zones. Sedimentation, eogenesis, fracturing, telogenesis, and compaction can be sequenced to simulate the inheritance effects of one process on another and quantify the impact on properties. This comprehensive approach to geological forward modeling of reservoirs delivers the evolution of petrophysical variables from the original deposit to the current state.

The move towards multi-process geological forward reservoir simulation suggests that it should be able to meet and overcome the new technical challenges associated with the energy transition, namely the data scarcity, resource reduction, and the multi-physics use of models. Such a workflow could then become an essential methodology for building models for fossil resources, geothermics, and CCS.

Biography

Gerard J. Massonnat is a TotalEnergies R&D Fellow, presently International Expert for Reservoir Geology and Geomodeling. After he obtained a master’s degree in Geology, he received his Ph.D. in Hydrogeology of fracture reservoirs and then graduated from IFP School in Petroleum Engineering and Project Development.  Starting as a reservoir engineer, he was successively in charge of field monitoring and field appraisal for offshore and onshore oil fields and became team leader for multidisciplinary studies. Then he held multiple positions in R&D, from managing integrated projects on carbonate reservoirs, petrophysical synthesis, geostatistics, and smart up-scaling. His work now focuses on the development of the next generation of modeling tools for both matrix and dual porosity reservoirs. Gerard has authored more than 130 technical communications in international meetings and journals and applied for 42 patents.  He served in several conferences and workshops and was a Distinguished Lecturer for SPE.