Fuel Processors for LPG

A breadboard laboratory fuel processor was then set up, which had an efficiency of between 56 and 60%. These values were lower than the calculations despite the 

The system was started by hydrogen combustion from ambient temperature, however, the source of this hydrogen was unclear for a practical system. Therefore, these workers planned to combust LPG in a future modification of the fuel processor. The start-up was performed in four steps. Firstly, the burner was started at 50% load and switched to 100% load after 2 min. At the same time, an air flow pre-heated the reactors of the fuel processor downstream of the reformer. After 16 min water was added to the reformer feed and after 21 min LPG was fed to the reformer at full load. At the same time, air was fed to the low temperature water-gas shift reactor to combust some of the hydrogen and carbon monoxide, similar to the procedure described by Springmann et al. [389] (see Section 5.4.5). The system then stabilised after 4 min, and therefore the total start-up time demand of the device amounted to 25 min. 

The preferential oxidation reactor failed to convert carbon monoxide to a level of less than 2000 ppm and thus connection of the system to the fuel cell was postponed. 

A completely different system, which worked with cracking of liquefied petroleum gas was presented by Ledjeff-Hey et al. [79]. The system had the charm of being, at first glance, much simpler compared with conventional reforming coupled with catalytic carbon monoxide clean-up. It was very similar to the concept presented 

---