Uncertainty evaluation and model optimization in multi-source reservoir modeling
Abstract The incompleteness of data and the variability in research methodologies can lead to significant uncertainties in reservoir modeling predictions. Effectively reducing and assessing these uncertainties are central issues in reservoir modeling studies. This research takes the ultra-low permea...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-05-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-04185-z |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract The incompleteness of data and the variability in research methodologies can lead to significant uncertainties in reservoir modeling predictions. Effectively reducing and assessing these uncertainties are central issues in reservoir modeling studies. This research takes the ultra-low permeability and ultra-low porosity conglomeratic sandstone oil reservoir in the Yong 935–936 block of the Dongying Depression as a case study to conduct a survey on uncertainty modeling and evaluating. Considering the presence of multiple sediment sources in the study area, a variogram model with local variable azimuth angles is designed to address the discontinuity of sand bodies at the partition boundaries during modeling. Geological insights are used to delineate the boundaries of fan bodies, and statistical analysis of the proportions of sandstone and mudstone under different effective reservoir property limits within these fan bodies provides conditional data to reduce uncertainties in the modeling process. A full-factor experiment is conducted to perform a sensitivity analysis on parameters influencing reserves calculations, clarifying the significance of various influencing factors, such as fan body boundaries, effective reservoir property thresholds, and variogram range. A multivariate regression model between reserves calculation and significant influencing factors is constructed through response surface experiments. Finally, combining the Monte Carlo random simulation method, the distribution of cumulative probability reserves is obtained, and the 3P reserves of the study area are predicted to test the effectiveness of the multivariate regression model. |
|---|---|
| ISSN: | 2045-2322 |