Well-testing based turbidite lobes modeling using the ensemble smoother with multiple data assimilation

Abstract

The representation of geological bodies is a difficult task, which involves a large number of parameters and assumptions that are commonly simplified in object-based modeling. A famous method that has been extensively applied for modeling geological bodies is the use of non-uniform rational B-Spline curves (NURBS) to delimit the boundaries of an object. Although NURBS provides highly detailed models, it only considers geological observations and assumptions. Moreover, the use of NURBS neglects information obtained from well-testing. This method also requires the complex and time-consuming process of determining the interior of the object in the parametric space. This is a classic problem in computational geometry, known as point location. To address these problems, this study proposes a well-testing-based object-based model of turbidite lobes using the ensemble smoother with multiple data assimilation. To escape the point location problem, we use single-valued B-Spline curves (SVBS) to build the turbidite system model. These curves are planar type B-Spline curves but they are defined as functions. The use of SVBS avoids the use of all complex algorithms and structures for solving the point location problem in the parametric space, if it is straightforward to decide if a point is in or out of the object. Consequently, it is possible to use the ensemble smoother with multiple data assimilation method to estimate the geometric parameters of the object, where a large number of realizations is required.