| Zeit: |
21. Juli 2011 |
| Referent*in: |
Meisam Ashraf
SINTEF,Oslo |
| Veranstaltungsort: |
Pfaffenwaldring 61, Raum U1.003 (MML), Universität Stuttgart
|
| Download als iCal: |
|
Geological storage of CO2 is a proposed solution for fighting the climate changebelieved by some to be a crucial issue of the century. The controversial discussions on the topic requires a scientific basis to demonstrate the feasibility of theprojects to the society, assessing the possible risks existing in the operation.Thelarge time and space scales in the process, accompanied with huge geologicaluncertainties emphasizes the need for sensitivity analysis on the model to findthe significance of the playing parameters. This is essential for studying thefeasibility of the operations and is considered to be a directive to plan investingfor uncertainty reductions.In this work, we study the uncertainty impact of geological modeling on theinjection and early migration of CO2 into underground aquifers. We use theset of SAIGUP geological models and run a scenario of injection on them. Toimprove the run speed dramatically, and keep the objectivity in the modelinguncertainty on its lower limit, we use arbitrary chaos expansion in a polynomialspace. Here the fundamental governing system of equations are approximated bypolynomials expanded in a space which is spanned by orthogonal basis. Thesebasis are constructed from arbitrary uncertainty distributions in the input pa-rameters. For global sensitivity analysis we use Sobol indices which are robustfor a general problem, regardless on the non-linearity and complexity of themodel. We conclude the results by a Monte-Carlo risk analysis using our fastpolynomial model.
