3D Multiple-points Statistics Simulations of the Roussillon Continental Pliocene Reservoir Using DeeSse
V. Dall'Alba, P. Renard, J. Straubhaar, B. Issautier, C. Duvail and Y. Caballero
Event name: Petroleum Geostatistics 2019
Session: Poster Session 1
Publication date: 02 September 2019
Info: Extended abstract, PDF ( 668.27Kb )
Price: € 20
This study presents a novel workflow that was developed to model the internal heterogeneity of a complex 3D reservoir using the Multiple-point Statistics (MPS) algorithm DeeSse. We propose to demonstrate the applicability of multivariate MPS simulation on a complex study site in the south of France. The modelled reservoir is the Continental Pliocene layer (PC) that is part of the Roussillon reservoir in the Perpignan's region. For this purpose, we use the direct sampling algorithm DeeSse and demonstrate its applicability on a large study site. New procedures are proposed to account for known geological constraints during simulations. In order to represent the complex sedimentary history of the plain, we create a non-stationnary training image (TI) that is used coupled to auxiliary variables maps during the simulation.