Integrating design of experiments and Bayesian model selection for the influence identification of geological parameters on the GAGD process in a multilayer heterogeneous sandstone oil reservoir
Document Type
Conference Proceeding
Publication Date
1-1-2017
Abstract
The Bayesian Model Selection (BMS) was adopted as a stochastic linear modeling to determine the most influential geological parameters affecting the Gas-Assisted Gravity Drainage (GAGD) Process performance in a multilayer heterogeneous sandstone oil reservoir/South Rumaila field. In the GAGD process simulation, CO2 is injected through vertical injectors at the top two reservoir's layers. Horizontal oil producers were set up through the layers of the highest oil saturation. Horizontal permeability, anisotropy ratio, and porosity were tested for their influence given each layer or group of layers. Latin Hypercube Design was adopted to create low-discrepancy experimental simulation jobs, which are then conducted to eliminate the non-influential factors through the linear BMS modeling. BMS adopts posterior probability to choose the best model, which has the optimal subset variables, among a set of candidate models. From the BMS results, accurate effect of each factor was investigated in details. Based on the concept of the GAGD process, it was generally noticed that permeability is more influencing than anisotropy ratio with no influence for porosity.
Publication Source (Journal or Book title)
Society of Petroleum Engineers - SPE/IATMI Asia Pacific Oil and Gas Conference and Exhibition 2017
Recommended Citation
Al-Mudhafar, W., & Rao, D. (2017). Integrating design of experiments and Bayesian model selection for the influence identification of geological parameters on the GAGD process in a multilayer heterogeneous sandstone oil reservoir. Society of Petroleum Engineers - SPE/IATMI Asia Pacific Oil and Gas Conference and Exhibition 2017, 2017-January https://doi.org/10.2118/187006-ms