Impact of Gas-Phase Reactions on SOFC Systems Operating on Diesel and Biomass-Derived Fuels

Abstract

The use of diesel fuel to power a solid oxide fuel cell (SOFC) presents several challenges. A major issue is deposit formation in either the external reformer, the anode channel, or within the SOFC anode itself. These deposits are generally poly-aromatic hydrocarbons (PAHs) produced either by gas-phase pyrolysis of the fuel or by catalytic reactions. In this report we describe n-hexane and ethylene pyrolysis experiments under conditions relevant to reformer or SOFC operation (τ=~1s, T=550~900°C, P~0.8 atm) to explore the potential for gas-phase reactions to produce deposit precursors. N-hexane is very reactive under these conditions and forms significant amounts of olefins (mainly ethylene) which can lead to deposits. The ethylene experiments also demonstrated that higher molecular weight species (deposit precursors) are rapidly formed. Under autothermal reforming conditions, such pyrolytic reactions are possible upstream of the catalyst bed if the fuel, air, and steam streams are not fully mixed. If part of the fuel does not mix with the oxidizer it will simply pyrolyze. At the same time, the remaining fuel fraction mixes with the entire oxidant inlet and thus creates higher local oxidant to fuel ratios than expected. Reaction of this leaner mixture can lead to temperature overshoots as more CO2 is formed. We have used a validated detailed kinetic model for ethane to explore the impact of incomplete fuel mixing on reforming performance. If only half the fuel mixes with the oxidants, this approach predicts formation of ethylene in the pyrolysis zone and excess CO2 with associated very high temperatures in the oxidation zone. This case could result in both excessive deposit formation as well as potential thermal damage to the downstream catalyst. On the other hand, assuming perfect fuel mixing under exothermic ATR conditions (τ=~1s, Ti=800°C, S/C=1.25, O/C=1.4), the gas phase reactions alone are sufficient to drive the system to equilibrium (no olefins or methane formed) due to the substantial increase in temperature. These results demonstrate the necessity for complete mixing of the fuel stream with the oxidant streams to limit both olefin production (and subsequent deposit formation) as well as the temperature overshoots. The model predictions for ethane as fuel suggest that the temperature should be kept below 500oC and the residence time in the mixing region should be minimized to avoid these undesired gas reactions. Since actual diesel fuel is expected to be even more reactive than ethane, the impact of gas-phase reactions is expected to be even greater than predicted in this study.