The usual workflow for delivering production forecasts for oil-bearing subsurface reservoirs consists in the building of very detailed geological and petrophysical models on which fluids flows are simulated. Populating models with heterogeneous distributions of reservoir properties is a very challenging issue because of the lack of data in typical datasets, that comprise boreholes with sparse lateral sampling, and seismic data with low vertical resolution.
Against this background of limited data, reservoir analogues have long been used to improve understanding of subsurface fields. From outcrops, geological data have been collected. Their interpretation has been used to build conceptual static models, and to run sensitivity studies of dynamic response to geological and petrophysical heterogeneity. Although quite advanced, this linear workflow does not achieve the objective of reservoir characterization as there is not any hydraulic information for calibration of the outcrop interpretation.
In the same time, hydrogeologists have developed observatories, dedicated to the observation and monitoring of groundwater reservoirs, for the exploitation of water resources and the prevention of environmental risks. These experimental sites provide data relevant to the understanding of the water cycle and the behavior of the aquifers. However, despite the acquisition of geological, geophysical and petrophysical data in boreholes, the description of geology is usually lacking in detail, thereby compromising the quality of the coupling between measurements, theories, and models.
A new concept of dynamic outcrop analogue arose from a double need: on one side, the integration of hydrodynamics and geophysics in outcrops studies, on the other side the use of detailed geology in observatories studies. The dynamic outcrop analogue allows the characterization of the reservoir at different scales and from an integrated point of view. The analogue is assessed through multi-physics experiments on various supports : outcrops, but also plugs, cores, boreholes and between wells, caves, tunnels... Coupling the hydrodynamic response of groundwater reservoir to a detailed static description of its heterogeneity very early before the building of models provides a key knowledge on the understanding of reservoir architecture. It contributes to improve the way models are populated and to enhance the forecast of reservoir.
This new methodology has been developed and tested through a pilot project, ALBION. The demonstration project is dedicated to the Urgonian carbonates (a barremo-aptian series outcropping in South-East France) and proposes a multi-scales set of analogues in the Fontaine de Vaucluse catchment area. On other sites, further projects are dedicated to various topics: multi-scales permeability in tidal sandstones, anisotropy of permeability in anastomosed fluviatile channels, heterogeneity in unconsolidated alluvial coarse gravels, fractures and karstic networks, and permeability in a fault zone.
The authors thank TOTAL for funding this R&D project and giving permission to present this work.