Coupled Hydro-Geomechanical Modelling of the Cold Production Process

Abstract

Abstract Aspects of the cold production process have been studied mechanistically employing a unified framework and coupled geomechanical-fluid flow simulation models. The unified framework is a generalization of that proposed by Papamichos et al1 such that stress concentration causes mechanical weakening, and fluid flow causes erosional mobilization of sand particles. Porosity change is the primary coupling parameter whereby the soil mechanical and strength properties are altered as a continuum damage process, while erosional generation of mobilized sand particles causes fluid porosity increase and oil-sand slurry flow. These ideas are implemented in coupled simulation models using different coupling strategies and different physical assumptions to study various aspects of the cold production process. The first coupled model is a stand-alone finite element model for two-dimensional radial plane strain and single phase fluid flow. Here the geomechanics and fluid flow aspects are fully coupled such that the time dependent (undrained to drained) mechanical stress transfer is strictly preserved. The second coupled model allows a more complete fluid model (including gas ex-solution and water influx), as well as a time dependent sand mobilization mechanism, but at the expense of a weaker (time explicit) geomechanical-fluid flow coupling, through Settari's concept of volume (effective porosity) coupling21. These models are applied to simulations of cold production at the laboratory and field scale. The laboratory scale models are based on the CT experiments of Tremblay et al2, which emphasize the erosional aspects of the cold production process, and its coupling to gas exsolution. The field scale models are based on single well production observations in Alberta bitumen reservoirs, such as Yeung et al44 and include sensitivities to water influx. Additional geomechanical aspects which may be important in deeper reservoirs such as those in Venezuela are also illustrated. Finally implications and recommendations for future simulation modelling at the field scale are provided.