Regensburg 2016 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
DY: Fachverband Dynamik und Statistische Physik
DY 30: Poster - Complex nonlinear systems
DY 30.15: Poster
Dienstag, 8. März 2016, 18:15–21:00, Poster C
Application of reduced-order model based on cluster analysis in numerical simulation of low permeability reservoir — •Weiwei Li1,2, Yanyu Zhang1, and Jianghai Lv3 — 1Saarland University, Experimental Physics, D-66041 Saarbrücken, Germany — 2College of Petroleum Engineering, China University of Petroleum, 266580, Qingdao, China — 3PetroChina ChangQing Oilfield Company Oil Production No.5, 710200, Xi'an, China
A mathematical model for a two dimensional reservoir considering the influence of start-up pressure gradients is solved based on a fully implicit finite difference scheme. The reduced-order model is established by projecting the original model on the low dimension which is formed by the basis function produced from the collected snapshots. The Lloyd algorithm for centroidal Voronoi tessellation is employed for cluster analysis to tackle the potential shortcomings from collecting snapshots with same time intervals. The computational results show that the used reduced-order model based on proper orthogonal decomposition is able to effectively approximate the results of the original model. With the help of the cluster analysis, data reduction is achieved on one hand, which could create uniform snapshots in space resulting to the enhanced accuracy of reduce-order model. On the other hand, extremely limited clustering groups might lead to loss of information and thus impairing the accuracy of the final reduced-order model.