A Dynamic Control-Volume Data-Reconciliation Framework for Transient Methane Pyrolysis in a Thermal Plasma Reactor

Transient methane pyrolysis in a DC thermal plasma reactor was assessed using a zero-dimensional control-volume reconstruction and elemental balance consistency framework. Programmed Ar and CNG feed rates, time-aligned outlet gas readings, a conditional Ar cross-sensitivity correction for the H2 detector, and finite gas holdup were combined with C/H/O residuals, a lumped thermal state, and exploratory kinetic and exergy scenarios. For the principal 10–25 min interval, the nominal conditional CH4 conversion is 94.2%; the 93.2–95.1% Monte Carlo interval propagates selected nominal analyzer specifications only and is not an experimental confidence interval because the raw H2 readings (approximately 51–59 vol.%; the corrected mean is 59.2 vol.%) exceeded the listed 0–10 vol.% range. Deterministic ±20% H2-channel gain scenarios give 91.95–95.46% conversion but are not calibration-error bounds. A hypothetical air-ingress scenario that closes the mean oxygen balance yields 93.84% conversion with a remaining hydrogen residual of 13.35 mol-H min−1; C2Hγ scenarios do not simultaneously reconcile the hydrogen balance and reported recovered carbonaceous mass. The dynamic and quasi-steady outlet-flow reconstructions nearly coincide (mean ratio: 1.0001; effective gas residence time: ≈0.25 s versus an estimated 1000 s lumped thermal scale). The 4–34 min integrated carbon-equivalent residual is 2.31 kg-C, compared with 1.875 kg of reported recovered carbonaceous material of unverified purity and run attribution. Measured H2 is approximately 49% of the ideal CH4-to-C(s)-plus-2H2 yield, with no unique explanation for the hydrogen deficit. The oxygen residual (−0.143 mol-O/min) is approximately 11.4 times the oxygen input from the stated CNG feed and is therefore not a satisfactory oxygen closure. Reduced kinetic and exergy calculations remain exploratory, assumption-dependent diagnostics, not independently validated performance predictions.