Multistage Preconditioner for Well Groups and Automatic Differentiation for Next Generation GPRS
Abstract/Contents
- Abstract
- This report includes two distinct parts: "Multistage Preconditioner for Well Groups" and "Automatic Differentiation for Next Generation GPRS". The first part describes the following work: (1) Extension of the CPR-based multistage preconditioner to group constraints on multi-segment (MS) wells; (2) modification of the data structure for adaptively working with the AIM (Adaptive Implicit Method) formulation and well-control switching; and (3) estimation of initial junction pressure for group-rate control and junction-pressure control mode with initial-rate estimates. The extensions are tested with large homogeneous and heterogeneous cases with group constraints. With the above improvements, GPRS is able to solve fully coupled reservoir-facilities simulation including well groups, with either FIM (Fully Implicit Method) or AIM, in an efficient and accurate manner. The second part describes two extensions of the ADETL (Automatically Differentiable Expression Templates Library) framework, which was conceived and designed by Rami Younis. Three primary stages (building, evaluation, and tearing down) are involved in the automatic generation of gradients behind the scenes. The extensions performed as part of this research effort include: (1) three new allocator options for the expression building stage, and (2) a block-sparse data structure for the expression evaluation stage. Additional improvements have also been implemented and integrated into ADETL. A variety of numerical examples are investigated and comparisons are made between GPRS and ADETL-based simulators. With the new allocator and block-sparse data structure, the overhead of ADETL ranges from -4% to 45%. Automatic differentiation is a flexible and efficient approach for building complex reservoir simulators. The next-generation of GPRS will be based on ADETL.
Description
| Type of resource | text |
|---|---|
| Date created | June 2009 |
Creators/Contributors
| Author | Zhou, Yifan |
|---|---|
| Primary advisor | Tchelepi, Hamdi |
| Primary advisor | Aziz, Khalid |
| Degree granting institution | Stanford University, Department of Energy Resources Engineering |
Subjects
| Subject | School of Earth Energy & Environmental Sciences |
|---|---|
| Genre | Thesis |
Bibliographic information
Access conditions
- Use and reproduction
- User agrees that, where applicable, content will not be used to identify or to otherwise infringe the privacy or confidentiality rights of individuals. Content distributed via the Stanford Digital Repository may be subject to additional license and use restrictions applied by the depositor.
Preferred citation
- Preferred Citation
- Zhou, Yifan. (2009). Multistage Preconditioner for Well Groups and Automatic Differentiation for Next Generation GPRS. Stanford Digital Repository. Available at: https://purl.stanford.edu/br498vc8942
Collection
Master's Theses, Doerr School of Sustainability
View other items in this collection in SearchWorksContact information
- Contact
- brannerlibrary@stanford.edu
Also listed in
Loading usage metrics...