Coalfield fires persist for years and spread through loose overburden where temperature rise, fracture opening, and oxygen supply interact. Tao et al. (2026), Fuel, combine a 3D physical similarity experiment with COMSOL Multiphysics thermal-fluid-solid coupling to describe how fracture networks sustain combustion via chimney-like oxygen transport. Below is a structured reading note—not a substitute for the full paper.

Paper highlights

  • Coalfield fire temperature fields show spatiotemporal asymmetry and hysteresis;
  • A self-reinforcing cycle forms: fracture expansion → oxygen supply → intensified combustion;
  • Fire-source O₂ can fall to 4.4 % (≈800 K center) while fracture channels hold 15–18 %;
  • Combustion is an unsteady thermal-fluid-solid process driven by convection, oxidation, and gas flow.

1. Problem background

1.1 Why coalfield fires are hard to control

Unlike confined goaf spontaneous combustion, surface/near-surface coalfield fires involve open or semi-open strata, long duration, and large spatial extent. Heat weakens rock/coal, opens fractures, and draws air downward—fueling oxidation even where the core is oxygen-starved.

1.2 Knowledge gap

Many studies treat temperature, gas, or mechanics separately. This work asks how fracture propagation and oxygen pathways co-evolve with the temperature field over time—the coupling needed for prevention zoning and injection/sealing design.

2. Methods

2.1 3D similarity experiment

The authors reproduce coalfield fire combustion on a 3D physical similarity simulation platform, preserving scaled geometry and boundary conditions so lab time maps to field-scale thermal and flow behavior.

2.2 COMSOL thermal-fluid-solid coupling

Numerical models in COMSOL Multiphysics couple:

| Process | Role in the model | |---------|-------------------| | Heat transfer / thermal stress | Drives thermal expansion and strength loss | | Solid mechanics / gravity | Captures instability and fracture growth | | Fluid flow / gas transport | Resolves O₂ supply and chimney effects | | Coal oxidation | Links temperature to reaction heat release |

Together these represent unsteady interaction among thermal convection, coal–oxygen reaction, and gas flow.

3. Key results

3.1 Temperature field evolution

  • The temperature field exhibits marked asymmetry and hysteresis (history-dependent lag between heating and cooling paths).
  • Maximum coal-seam temperature reaches 502.5 °C.
  • Roof peaks at 157.4 °C around 94 h; overlying strata reach 85.6 °C around 135 h—showing delayed upward heat penetration.

3.2 Fracture-driven feedback

Thermal stress and gravity-induced instability jointly drive fracture propagation, forming a penetrating fracture network. Fractures act as preferential air paths (chimney effect), boosting O₂ delivery to hot coal and closing a positive feedback loop:

Fracture opening → enhanced O₂ transport → stronger combustion → more thermal stress → further fracturing

3.3 Oxygen distribution paradox

Numerical results highlight a split oxygen landscape:

| Location | Approx. O₂ | Interpretation | |----------|------------|----------------| | High-temperature fire core (~800 K) | 4.4 % | Local consumption/depletion at the reaction front | | Fracture channels | 15–18 % | Persistent air supply corridors sustaining the fire |

Fracture channels therefore function as the dominant O₂ transmission network, even when the source zone is severely depleted—explaining why sealing surface cracks or isolating chimney paths is often more decisive than cooling the hot spot alone.

4. Engineering implications

  • Monitor fractures and air paths, not temperature alone: channel O₂ can keep combustion alive while core readings look anoxic.
  • Break the feedback cycle early—grouting, cover compaction, or barrier injection before penetrating networks form.
  • Asymmetric, hysteretic temperature maps imply single-point thermometry may mis-rank risk; favor spatial arrays and time-lapse interpretation.
  • Pair mechanism studies with gas (O₂, CO, CO₂) and surface deformation monitoring for operational early warning.
  • Similarity + FEM parameters must be recalibrated per coalfield lithology, dip, and weather-driven leakage.

5. Limits and reproduction notes

  1. Lab similarity scaling: transfer coefficients and fracture laws need site validation.
  2. Not an ML/forecast paper: value is in coupled physics and qualitative zoning of O₂ pathways.
  3. Weather and topography at real coalfields add seasonal leakage—not fully captured in controlled experiments.
  4. Prevention remains physical: models inform where to seal and inject; they do not replace field extinguishing operations.

Reference

Tao, F.; Ren, L.-F.; Wang, C.-P.; Song, Z.-Y.; Li, X.; Wang, J.-R.; Jia, Y.-Z.; Deng, J. Study on Spatiotemporal Evolution Characteristics of Temperature Field, Fracture Propagation and Oxygen Transport in Coalfield Fire. Fuel 2026, 411, 138034.