Construction of geostatistical aquifer models integrating dynamic flow and tracer data using inverse technique

摘要: Natural aquifers are heterogeneous, and geostatistical methods widely used to simulate the heterogeneity of aquifer properties. Due limited available data, it is essential integrate as much information possible reduce uncertainty in models flow predictions. Traditional techniques efficiently consider static hard soft information, such core data seismic data. However, dynamic transport rates, pressure, tracer breakthrough, important that not easily considered with traditional techniques. Integrating into a model requires solution difficult inverse problem, since properties related each other through non-linear equations. A recently developed geostatistically based technique, sequential self-calibration (SSC) method, introduced those The SSC method an iterative technique coupled optimization procedure. It provides for fast generation multiple realizations property jointly match pressure breakthrough yet display same characteristics. This flexible, computationally efficient, robust. main features include (1) master point concept reduces number parameters, (2) perturbation mechanism on kriging accounts spatial correlation properties, (3) streamline-based simulator integration (4) new semi-analytical computing sensitivity coefficients breakthrough. Applications demonstrated synthetic set. Results show carry variation permeability inter-well areas. As contrast, provide at near well-bore areas only. leads significant improvement representation reduction model. accuracy predictions can be dramatically improved by integrating