Weijiao Ma, Yunpeng Wang

Thermal decomposition of sedimentary organic matter may contribute to natural H2 generation, but measured H2 yields and kinetic parameters can be sensitive to experimental conditions. Nine coal and carbonaceous mudstone samples from the Kuqa Depression, Tarim Basin, were investigated using open-system pyrolysis–gas chromatography and closed-system pyrolysis. Open-system pyrolysis was conducted at heating rates of 0.5, 1, 2, and 5 °C/min and N2 purge rates of 5, 10, 15, and 25 mL/min. Closed-system experiments were performed at 0.25 and 2.5 °C/min on a representative sample. In the open-system experiments, CH4 was generally released as a single dominant peak, whereas H2 appeared later, over a broader and higher-temperature interval, and commonly exhibited two partially overlapping components at 0.5 °C/min. Increasing the heating rate shifted the peak temperatures upward, shortened the release intervals, and generally increased the cumulative yields. The first H2 component of KZ01 was deconvoluted for kinetic analysis. Results show that apparent activation energies vary with the choice of heating-rate combination. The slower and faster heating-rate combinations yielded dominant activation-energy ranges of 39–43 and 43–49 kcal/mol for CH4, whereas yielded 66–72 and 42–50 kcal/mol for the first H2. Extrapolation to a geological heating rate of 3 °C/Ma produced onset-temperature differences of approximately 16 °C for CH4 and 82 °C for H2, demonstrating the greater sensitivity of the H2 prediction to curve selection and peak deconvolution. Increasing N2 purge rate reduced the apparent gas-release intensity and cumulative yields without substantially shifting peak temperatures. Closed-system pyrolysis yields higher CH4 but substantially lower H2 than open-system experiments. The unusually high activation energies of 137–155 kcal/mol for H2 in closed-system pyrolysis may describe net free-H2 accumulation after simultaneous generation, consumption, and redistribution rather than intrinsic reaction kinetics. These findings demonstrate that experimentally determined H2 yields and kinetic parameters are strongly experimental condition-dependent, and that reliable geological prediction requires differentiation between gross H2 generation potential and net free H2 accumulation. Caution is urged in interlaboratory comparison and basin modelling applications of organic H2 kinetics.