Plasma-Assisted Partial Methane Oxidation. Part I: One-Dimensional Statistical Modeling of a Dielectric Barrier Discharge Reactor

This paper presents an original one-dimensional statistical model designed to complement experimental data from a practical dielectric barrier discharge (DBD) reactor. The experimental setup consists of a plasma-assisted reactor with gas injection composed solely of CH4 and O2. The numerical procedure uses electrical measurements to provide a realistic description of the power consumption and the current flowing through the gas, in a complex scenario where approximately one hundred current peaks are measured per electrical half-period. A plasma kinetic model is then used to analyze the chemistry and characteristic times of both isolated discharges and a representative train of discharges occurring within the reactor. These times are used to optimize the computational costs and to select the most appropriate kinetic schemes for the gas phase, whether in a plasma or quasi-thermodynamic equilibrium state. This approach also allows the separation of fast transformations (plasma) occurring at constant volume, from slower transformations occurring at constant pressure. The statistical approach, based on a Monte Carlo method, clearly identifies the assumptions required to reduce the real complexity of the DBD reactor to a 1D flow model. The combination of chromatographic measurements at the reactor outlet and numerical simulations provides the heterogeneity factor of the discharges, which is identified as a key parameter in the model. Although the flow can be considered stationary on average, the obtained value reveals a highly heterogeneous spatial distribution of the discharges within the reactor. Thus, the numerical results suggest that the gases passing through the reactor are rarely in a plasma state.