Quantitative modelling of fluvial sedimentary architecture and its application to evaluating suitability of subsurface reservoir successions for carbon sequestration

The creation of fluvial reservoir models that are geologically realistic remains challenging. A workflow has been developed for modelling reservoir successions that comprise fluvial meander-belt deposits, based on algorithms that employ multi-point statistics (MPS). A library of training images from which MPS modelling algorithms can borrow geological patterns for modelling meandering fluvial systems has been built. The training images incorporate sedimentary architectures relating to various types of point-bar deposits as observed in high-sinuosity river systems and their preserved deposits in the geologic record. The training images are applied to two widely employed MPS modelling algorithms: SNESIM and DEESSE. Solutions to common issues encountered in MPS modelling workflows have been established through optimisation of modelling settings.

Facies models that incorporate different levels of heterogeneities for meandering fluvial systems have been produced. The created facies models are used to simulate the injection of a CO2 plume applied into a carbon capture and storage (CCS) process, allowing a comparison between macroscale 3-facies models and mesoscale levels of heterogeneities as 5-facies models, incorporating intra-point-bar stratigraphic features. An analogue database (PAFD, Petrophysical Analogue Fluvial Database) is developed including petrophysical data related to fluvial successions. PAFD is used to support property static modelling, whose outputs are applied in property models. With more than 4,000 records, PAFD can be employed to inform subsurface modelling in data-poor situations.

The study demonstrates how CO2 dynamic simulations related to a CCS context are influenced by the underlying facies framework in a reservoir model. Important implications regarding the redistribution of pressures in the reservoir, caprock pressure relief phenomena, horizontal distribution and differences on capillary trapping mechanisms only emerge from models that consider meso-scale features. Furthermore, geological realism in terms of an accurate facies model proved to be fundamental in controlling the CO2 plume displacement, injection rates and the cumulative injected volumes.