Restarting a History-Matching Study for an Existing Reservoir Model

C02 RESTARTING A HISTORY-MATCHING STUDY FOR AN EXISTING RESERVOIR MODEL Abstract 1 Sometimes reservoir engineers want to update a previously built reservoir model to integrate newly collected data. Unfortunately the data used to generate the existing reservoir model are often lost. To avoid restarting the history-matching study from the very beginning we developed several techniques to infer the missing data. These data consist of a random function characterized by geostatistical parameters and a set of random numbers. The existing reservoir model is the realization associated to the estimated random numbers given the estimated random function. Once the random function and