A Study on System Identification and Input Design for SAGD Reservoirs

• Author / Creator

  Yao, Song

• Physics-based large-scale reservoir models are routinely employed in the prediction of the SAGD (Steam Assisted Gravity Drainage) process under different operation situations. However, due to the uncertainty of the reservoir and the limitations of the commercial reservoir simulators, the computational time is highly associate with impractical simulated results for all locations, especially when uncertainty and unexpected operational parameters are included. This thesis develops a system identification proxy model to forecast the SAGD reservoir production. Several combinations of system identification model structures and input datasets are tested for short-term predictions to understand the impact of model structures and input selection on the proxy model performance. Then recursive proxy model estimations are performed to increase the accuracy for long-term production prediction. After trying to improve the model fits for a multiple-step proxy model prediction, a set of input design cases using a simulator-built model are re-identified to test the validity of open-loop simulation by a data-driven proxy model. The data-driven proxy model could be used in reservoir management and optimization or to reduce the computing load.

• Subjects / Keywords

  • SAGD
  • System Identification

• Graduation date

  Spring 2014

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3S756T4G

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.