Reservoir Characterization and Prediction Uncertainty Assessment through Scale Splitting

1
Reservoir Characterization and Prediction
Uncertainty Assessment through Scale Splitting

• Trond Mannseth

RF-Rogaland Research
CIPR University of Bergen
2
Background

• Predict reservoir flows

Physical/Numerical model
Property model
Calibrated property model
Calibrated permeability model
3
Work Flow - Alternatives
Need good initial guess

• CPU intensive

Insufficient information
Selection of zones?
Finer scale variations?
4
Selection of Zones Multiscale Estimation
Ordinary ME

Reference permeability

Adaptive ME
5
AME Some Examples
6

7

8

AME Predictor Level Set Corrector
9
Incorporating Fine Scale Variation
Uncertainty assessment
Fine scale history match

10
Uncertainty Assessment

11
Project Summary
Time frame 1/7 2002 30/6 2005
Budget 8200 (5400) kNOK
Industry partners ENI (Agip) Norsk Hydro
Collaboration CIPR-UoB HWU TAMU
Personnel A-A Grimstad T Mannseth R
Valestrand (RF) I
Berre H Kruger MØ Lien LK Nielsen (UoB)