Experimentally Constrained Dynamic Permeability Modeling of Commingled Production in Stacked Coalbed Methane Reservoirs: A GP-2 Case Study

Stacked coalbed methane (CBM) reservoirs can increase the drainage thickness of a single well, but commingled production is influenced by stress-sensitive permeability, gas desorption, water drainage, and interlayer heterogeneity. This study presents a three-segment reservoir model for well GP-2 in the Tucheng block. Pore-fracture compressibility was estimated from overburden low-field nuclear magnetic resonance measurements and compared with stress-dependent permeability obtained by the pulse-decay method. The resulting coefficient was used in the dynamic permeability relationship and held fixed during history matching. The model was calibrated against gas- and water-production data from the first 240 d. One-factor simulations were then run over a common 3000 d calculation window to compare the relative responses to geological, adsorption, and stimulation parameters. In the GP-2 base model, average gas rate increased with equivalent coal thickness, gas content, Langmuir pressure, stimulated area, and stimulated-region permeability; inverse responses were obtained for cleat-fracture porosity, proportional three-layer initial permeability, initial reservoir pressure, and Langmuir volume. Adsorption time and interlayer spacing had comparatively small effects. These trends are specific to the selected model and parameter ranges and should not be interpreted as validated long-term forecasts or established causal relationships. This study demonstrates a practical way to carry a laboratory-derived stress-sensitivity parameter into a multilayer field model.