History matching is an iterative process that modifies a reservoir model to reproduce field behavior. Due to the scarcity of data the true distribution of facies, porosity and permeability between widely spaced wells is unknown, resulting in high uncertainty.
As the spatial patterns of permeability...

History matching is an iterative process that modifies a reservoir model to reproduce field behavior. Due to the scarcity of data the true distribution of facies, porosity and permeability between widely spaced wells is unknown, resulting in high uncertainty.
As the spatial patterns of permeability and porosity often significantly affect the flow response, evaluating heterogeneity is key in history matching. This work presents a stochastic method for use in probabilistic and iterative history matching. We select a set of the best geostatistical realizations and reproduce their spatial patterns in subsequent iterations to create a new set of better-matched models. All models, throughout the process, honor well log data and continuity, modeled with the variograms. This paper uses four approaches: two using a global method and two using a regional method. Both methods improved dynamic history data matching while honoring all well data but each is suited to different times in exploration and production. We present the global method as a simple tool to improve models. Characterized as an update of the entire reservoir, it is useful when the number of wells and the dynamic history data are scarce. The regional method is more efficient to process large amounts of information, enabling the independent match of dynamic well data, avoiding mismatches with other wells

Fechado