D-EGR WGS Catalyst Development and Optimization, 03-R8326
Gordon J.J. Bartley
Inclusive Dates: 07/01/12 – 07/01/13
Background — SwRI has been actively developing its dedicated exhaust gas recirculation (D-EGR) concept and system within the HEDGE Consortium. A key to successful application is the amount of hydrogen (H2) that can be efficiently produced and fed back into the intake mixture. Rich operation of the D-EGR cylinder produces a significant amount of H2, but more is beneficial. The D-EGR cylinder exhaust also contains substantial amounts of carbon monoxide (CO) and water (H2O). A water gas shift (WGS) catalyst reacts CO with H2O to form H2 and carbon dioxide (CO2), but no WGS catalyst had ever been developed for this application or environment. Catalysts that SwRI had used to evaluate the concept were traditional three-way (TWC) exhaust formulations that achieved about 45 percent H2 production efficiency with minimal durability. If this efficiency could be increased to 70 percent and durability improved, an additional 2 to 3 percent brake thermal efficiency (BTE) is possible, a very significant technological advance.
Approach — Three different catalysts received from catalyst companies had been evaluated on the D-EGR engine. The one that provided the best performance became the reference formulation for this work. The catalyst was analyzed to obtain the overall elemental composition of the catalyst washcoat. Starting with this reference formulation, a matrix of 45 varying formulations was prepared on core samples for testing. SwRI's Universal Synthetic Gas Reactor® (USGR®) was used to perform the testing. Each catalyst’s WGS activity was evaluated over a fixed set of test conditions. Sensitivities to individual independent and dependent variables were used in a statistical approach to identify the direction of optimum formulation for WGS reactivity in the anticipated temperature regions. The optimum operating conditions for that formulation were extracted from the data.
Accomplishments — The project work is complete. Four formulations were identified as producing high levels of H2, and one in particular was isolated that produced a very high level. The efficiency was somewhere between 65 and 85 percent, depending on the calculation method and H2 measurement technique. Two key findings were that rhodium and barium both had a positive effect on H2 production, and there was a beneficial synergy between the two. This formulation was prepared on full-size substrates and tested under D-EGR conditions on a bench engine. The H2 production exceeded the 7-percent target, achieving a maximum of 8.35 percent. Unfortunately, the catalyst could not maintain this performance over the long term. In just a matter of hours, the activity, as measured by CO conversion efficiency, fell significantly. It is believed that continuous rich (reducing) operation results in coke build up in the catalyst, reducing the activity. The project was completed before any solutions to this issue were evaluated. One such option would be to operate the catalyst over a perturbated regime with one cycle in eight running lean (oxidizing). The hypothesis is that the lean cycle could oxidize the coke precursors before they have a chance to build up into coke deposits. Further work is needed to address this issue.