Detection of Preferential Flow Paths in Naturally Fractured Reservoirs: Application To Adaptive Upscaling

@inproceedings{IfrahRM2007,
 abstract = { Open fractures act as conduits for fluid flow, due to their high permeability and transmissibility. However, by their properties and geometry, all fractures will not impact the flow with the same magnitude. Hierarchical methods have therefore been developed for upscaling such systems. The principle is to represent differently fractures in the coarse-scale model according to their length relatively to the coarse gridblock size. Such a method is only precise for large fractures selection (contrary to the small fractures case), and can be inaccurate only considering fracture length as important parameter. Moreover, the preferential flow paths within a fracture network are usually not evaluated to select the main fractures that need to be accurately modelled. Automatic detection techniques of such paths are thus required. In this article, we present a method implemented to determine lower-cost paths within a graph, based on a connectivity list representation of the fractured medium. Both fractures and matrix are represented in the network model. Flow paths are computed from an injector well to a producer. Moreover, paths are characterized by a cumulated cost of their constitutive nodes at the producer well. Three criteria have been chosen to selection the main flow paths: the main conductive pipes linking two nodes, or the steady state two-phase (water-oil) flow rate characterizing a connection. The influence of the computed preferential flow paths on the flow simulation results is then evaluated by performing an adaptive upscaling which integrates these flow paths. },
 author = { Ifrah, Gabrielle AND Vitel, Sarah AND Caumon, Guillaume },
 booktitle = { 27th gOcad Meeting },
 month = { "june" },
 publisher = { ASGA },
 title = { Detection of Preferential Flow Paths in Naturally Fractured Reservoirs: Application To Adaptive Upscaling },
 year = { 2007 }
}