Initial Sampling of Ensemble for Steam-Assisted-Gravity-Drainage-Reservoir History Matching
Bibliographic record
Abstract
Summary History matching of a steam-assisted-gravity-drainage (SAGD) reservoir requires a large ensemble size for proper uncertainty assessment, which ultimately results in high computational cost. Therefore, it is necessary to reduce the number of realizations for SAGD-reservoir simulation purposes. In this paper, a novel sampling method (based on the probability-distance-minimization method) to generate an initial ensemble of reduced size is discussed. This method considers multiple static measurements and geological properties and uses Kantorovich distance to quantify the probability distance between the original ensemble and the reduced ensemble, which is later optimized by use of the mixed-integer linear-optimization (MILP) technique. To show the effectiveness of the method, we have shown history matching of an SAGD reservoir using the smaller size initial ensemble derived from the proposed method and compared with the original ensemble. For history matching, the ensemble Kalman filter (EnKF) has been used because of its ability to assimilate data for large-scale nonlinear systems. Results are compared with several other methods, such as importance sampling, kernel K-means clustering, and sampling by use of orthogonal ensemble members. The robustness and usefulness of each method for generating an improved initial ensemble of reduced size are analyzed on the basis of two criteria: (1) Does the smaller ensemble retain the same statistical distribution characteristics as the original ensemble, and (2) does the smaller ensemble improve the performance of history matching? In general, we conclude that the improved, smaller initial ensemble created by use of the proposed method retains the best statistical characteristics of the original ensemble. Also, it provides better performance compared with other ranking methods in sampling and history matching using EnKF. Finally, the proposed method can reduce the computing cost significantly without compromising uncertainty in the forecast model, which allows for real-time updating at smaller time intervals.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".