Many papers have described high-temperature steam reforming of ethanol at around 600 °C but...this temperature is too high to be driven by engine exhaust. Coking is also an issue because of the dehydration of ethanol at acid sites on the catalyst at high temperature, generating ethylene, which then carbonizes.

...A 2005 paper in this journal reported that ethanol could be reformed by a different pathway at about 300 °C, enabling the reaction to be sustained by exhaust heat. Unlike high-temperature reforming, supplementary water is not required. Thus, the reaction is compatible with anhydrous ethanol fuels, such as E85, without requiring a secondary water tank.

...Despite its much lower hydrogen content, research in single-cylinder engines over the last four years has demonstrated that the H₂/CO/CH₄ mixture, “ethanol reformate”, enables engine efficiency improvements comparable to those seen in reformed methanol engines. Further work revealed that, because of the enhanced flammability range of hydrogen, only 25-50% of the fuel needed to be reformed to obtain the efficiency benefits of highly dilute engine operation. This finding, along with the observation that reformate was not required at high engine load, implies that a reformer of modest size could be developed for automotive applications. Unlike reformers for fuel cell vehicles, low-temperature ethanol reformers need only provide a fraction of the fuel over the low and midload portions of the drivecycle to achieve significant fuel economy and emission benefits.

—Sall et al.

High-temperature reformers need to burn a portion of the fuel to maintain their reforming temperatures, thereby generating a fuel-economy penalty. Because exhaust-driven reforming uses only waste engine heat, it avoids this efficiency hit, the authors note. However, they add, this benefit comes at the cost of a substantial design complication: the reformer must serve as both a reactor and a heat exchanger between the fuel and the exhaust stream. Other design considerations for automotive applications include:

• The device must exhibit low thermal mass to reach operating temperatures early in the drive cycle—a challenge when there must be enough surface area to deliver adequate heat exchange while also ensuring a pressure rating sufficient to withstand the fuel-side backpressure created by the catalyst bed.

• The device must minimize exhaust backpressure to negligible values (inches of water) because exhaust backpressure places a load on the engine and reduces efficiency.

• A low-cost, low-weight, compact design capable of being packaged in a vehicle exhaust train.

The team developed three low-temperature ethanol reformer architectures (representing a design evolution) and tested them at automotive scale with exhaust from a V8 engine: shell-and-tube; finned tube; and “shoebox” reformer.

The linear finned tube reformers ultimately failed because of backpressure resulting from slow migration of the powder catalyst and the inexorable rise in backpressure that resulted.

Only the shoebox unit—consisting of transverse shell-and-tube design in which banks of parallel, vertical catalyst tubes extended through a transverse stack of exhaust-side heat-exchange plates—exhibited sustained high conversion with low and stable fuel-side pressure throughout a 500 h test period. This design has relatively low thermal mass and can be readily packaged on a vehicle.

The ethanol reformate produced by the copper/nickel sponge catalyst has been shown in earlier studies to enable very low emissions at cold start along with substantial improvements in fuel economy. At steady state, the shoebox reformer provides sufficient reformate to realize these benefits in a passenger vehicle. E85 and hydrous ethanol motor fuels are now widely available in the U.S. and Brazil, respectively, making this approach practical in these two countries, at least for vehicles with longer drive cycles, such as taxicabs or delivery vehicles, for which reformer heating time is less of a concern.

For an ordinary passenger vehicle with relatively short drivecycles, reduced heating time is clearly necessary. For experimental convenience, relatively long and massive exhaust lines connected the engine to the reformers in this study, resulting in heating times approaching 20 min. Even with substantial reductions in exhaust line thermal mass, downsizing of the reformer is likely to be required to achieve heating times significantly below 5 min. The excellent heat transfer and conversion observed in the shoebox reformers, particularly reformer H with reduced catalyst volume, make us confident that this can be achieved, although closer coupling to the engine and catalytic converters will likely be required. Experimental work in this direction is ongoing.

—Sall et al.

Resources

• Erik D. Sall, David A. Morgenstern, James P. Fornango, James W. Taylor, Nichilos Chomic, and Jennifer Wheeler (2013) Reforming of Ethanol with Exhaust Heat at Automotive Scale Energy & Fuels doi: 10.1021/ef4011274