Image-informed prediction of effective thermal conductivity in gas hydrate sediments using integrated reconstruction and meso-scale simulation

Effective thermal conductivity (ETC) is a key input for reservoir-scale temperature-field simulation, thermal stimulation design, and thermal safety assessment during gas hydrate exploitation. However, ETC in gas hydrate sediments (GHS) is highly sensitive to hydrate occurrence morphology, phase proportions, and structural heterogeneity, which limits the reliability of conventional correlation-based and idealized theoretical models. This study develops an image-informed integrated reconstruction and meso-scale simulation (IRMS) framework for mechanism-oriented and relative ETC assessment in GHS. The framework combines image-derived feature identification, digital structure reconstruction, steady-state heat-conduction analysis, and Monte Carlo sampling. A heat-conduction-guided hydrate discrimination strategy is introduced to identify thermally favorable regions for bridging and coating hydrates, so that hydrate placement better reflects physically plausible pore-scale evolution while improving computational efficiency. Finite-element simulations are then conducted for four representative hydrate occurrence patterns (bridging, coating, floating, and patch) to quantify ETC. The framework is evaluated against classical upper and lower bounds, a random reconstruction model, and published experimental data. The results show that gas content is the dominant negative control on ETC because of the low thermal conductivity of methane gas, while hydrate occurrence morphology further regulates ETC by reorganizing conductive pathways under the same phase fractions. The governing mechanism is the topology of conductive pathways, especially their continuity, width, number, and tortuosity. Floating hydrates are most beneficial at low solid contents because they create multi-path conductive networks, whereas bridging hydrates become more effective at high solid contents by strengthening interparticle thermal contacts. Particle anisotropy further induces directional thermal responses when grain long axes align with the imposed heat flux. The proposed IRMS framework establishes a quantitative link between hydrate occurrence characteristics and macroscopic thermal behavior, and provides an engineering-oriented workflow for zone-wise ETC assignment in reservoir thermal models, temperature-field prediction, and thermal management in gas hydrate exploitation.