Formation damage is generally limited in the immediate near-wellbore region and needs a particular near-well flow modeling using fine gridblocks. However, near-well models are usually developed standalone and are decoupled from reservoir models. Using a standalone near-well model, which does not take into account production scenarios, cannot correctly predict well injectivity or productivity.
In this paper, we propose a new technique for the coupled modeling of the near-well flow model and the reservoir model in a simple and consistent way. In this new approach, data are exchanged and updated through boundary conditions for the near-well model and numerical productivity index (PI) (or skin factors) for the reservoir model. Examples show that this coupled modeling gives quite satisfying results.
It has been widely recognized that formation damage has a huge impact on well productivities. The economic impact of poor productivity of some wells has pushed toward significant efforts in recent years to study laboratory testing techniques and numerical modeling methods for predicting formation damage. Formation damage is generally limited in the immediate nearwell region and needs a particular near-well flow modeling. Apart from complex near-wellbore physics, this modeling requires also very fine gridblocks around the well, which is not at the same scale as reservoir flow simulation. Some models have been developed to study the near-wellbore formation damage (Minssieux et al. , Veerapen et al. , Ding et al. [3–4], Bedrikovetsky et al. , Moghadasi et al. , Civan ), but these models are mostly used for the modeling of formation damage around a single well and are decoupled from reservoir production scenarios.
On the other hand, fine grid system around a well such as local grid refinement or hybrid grid techniques is developed for the reservoir simulator to better describe near-well flow behavior in a full-field reservoir simulation. In general, fluid flow behaviors are not the same in the reservoir and in the near-well region. Formation damage is not the unique factor leading to use fine grid in the near-well region. High pressure gradient and radial flow behavior need a particular treatment around the well. Near-well heterogeneity has also a great impact on the well productivity. Local grid refinement and hybrid grid are more and more used to improve accuracy of well modeling. However, taking into account formation damage, which is generally at a scale from several centimeters to several meters, will make the near-well modeling much more difficult with very fine gridblocks in the reservoir simulator.
Simulating simultaneously near and far well flow behaviors arises several mathematical problems. Because the gridblocks around the wellbore are small, the computational stability requires small time steps and a greater number of Newton-Raphson iterations. It is obvious that this is not feasible if the solution for the full reservoir and the near-well region must be performed simultaneously. To improve the simulation efficiency, various methods have been studied in the literature. In particular, techniques of domain decomposition (Ewing et al. , Gaiffe ) and the windowing approach (Deimbacher et al. , Mlacnik and Heinemann ) have been studied to take into account fine gridblocks with small time steps in the near-well region and large gridblocks with large time steps in far well regions.