Cellular Automata Applied to Fault and Surface Related Diagenesis in Carbonate Reservoirs

Early and burial diagenesis are the main processes modifying carbonate rock matrix properties. These processes have been studied extensively and are well-known. Detailed petrographic studies allow geologists to understand the sequence of events that increase or decrease the pore size and connectivity of the reservoir rock types. However, 3D spatial prediction of diagenetic trends or geobodies remains challenging, mainly due to interactions with the pre-existing geology, and because fault and/or fractures, and complex fluids/rock reactions through time produce a very wide range of shapes and sizes.

Traditional variogram-based approaches used to populate rock properties in static reservoir models are not able to capture very complex geometries such as cave networks formed in karsted terrains. It is also very difficult to build 3D training images in order to obtain efficient multipoint statistic predictions for such systems. By contrast, forward modeling approaches are emerging as promising tools to better understand and assess the key parameters for specific fault or surface related diagenetic processes. We present here a new approach based on cellular automata, which uses very simple cell-based rules in order to reproduce trends and geometries.

As examples we present two case studies on fault and surface related processes. The first case will address the ability of cellular automata to predict complex karst sytems in an unconventional oil reservoir. Taking into account the geological architecture, the cellular automata is used to predict the distribution of karst breccias, cave networks or sinkholes. The second case study concerns dolomitisation in a Middle East carbonate reservoir. In this example we will emphasize the ability of the cellular automata approach to be conditioned to well data in order to better predict the spatial distribution of the dolomitised geobodies.

AAPG Search and Discovery Article #90142 © 2012 AAPG Annual Convention and Exhibition, April 22-25, 2012, Long Beach, California