Synapse
⌘+K
Synapse
PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
July 10, 2026Scientific ReportsOpen Access

Predicting CO2 injection profiles in heterogeneous reservoirs using a physics-aware deep learning framework

View Full Paper
Ask AI
Bookmark
Share

Authors

ZZZihao ZhengHSHaoxi ShiXLXintong Liu

Discussion

Loading...

Member takes

Overview

Randomized trial predicts CO2 injection profiles effectively in heterogeneous reservoirs, suggesting improved strategies for optimization.

Key Points

  • The aim is to develop a deep learning framework to accurately predict CO2 injection profiles in diverse reservoir conditions.
  • Generated a large dataset using the ECLIPSE simulator under various geological scenarios.
  • Developed a deep learning model using Bi-LSTM, self-attention, and FiLM for prediction.
  • Quantitative evaluations and ablation studies validated the model's performance across multiple injection regimes.
  • Achieved a mean absolute error (MAE) of 14.28 ± 0.41 m3/d and an R2 of 0.9915 ± 0.0024 over five independent runs.
  • The model outperformed conventional LSTM-based methods, indicating higher prediction accuracy.
  • Demonstrated stable performance under different injection scenarios, enhancing applicability in CO2-EOR and CCUS.

Cite This Study

Zheng et al. (2026) studied this question.

synapsesocial.com/papers/6a508d096eeac72a437a0c91https://doi.org/10.1038/s41598-026-60614-7
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1High-fidelity spatiotemporal evolution of CO₂ storage in saline aquifers: A physics-synergized deep learning framework for CO₂ plume prediction and super-resolution reconstruction2026
  2. 2Deep Learning Framework for Accurate Static and Dynamic Prediction of CO2 Enhanced Oil Recovery and Storage Capacity2024 · 16 citations
  3. 3CO2 injection-based enhanced methane recovery from carbonate gas reservoirs via deep learning2024 · 10 citations
  4. 4Data-Driven Predictions of CO2 EOR Numerical Studies Using Machine Learning in an Open-Source Framework2024 · 4 citations
  5. 5An Efficient Hybrid Deep Learning Framework for CO2 Plume Monitoring in Geological Sequestration2025