A research group at the Engine Research Center at the University of Wisconsin-Madison has demonstrated significant improvements in test-engine efficiency by burning two fuels instead of one.
The dual-fuel experiments took place in a single-cylinder version of a heavy-duty Caterpillar diesel truck engine at their lab that injects a well-mixed gasoline-air charge and then overlays it with multiple injections of diesel fuel. “We’ve seen 20 to 25% improvement in fuel-efficiency numbers and thermal efficiencies up to 60%, while meeting U.S. emissions standards for NOx and particulate in-cylinder [without aftertreatment],” said Rolf Reitz, the team leader and a principal investigator at the research center.
For comparison, the thermal efficiencies of diesel power plants using other low-temperature combustion strategies such as homogeneous charge compression ignition (HCCI), premixed charge compression ignition (PCCI), and modulated kinetics top out at around 45%. Since these schemes generally have problems at high loads and offer no direct control of combustion timing, engineers have been investigating “hybrid” approaches in recent years, Reitz said.
“The extra fuel adds another degree of freedom to the system,” he explained. “Essentially, it gives you another knob with which to control the combustion process. And because the combustion temperatures are low, we get fewer heat-transfer losses through the cylinder walls and engine exhaust.”
The Wisconsin engine physics and chemistry research group, which includes Wisconsin graduate students Reed Hanson, Derek Splitter, and Sage Kokjohn, relies on in-cylinder fuel blending using port fuel injection of gasoline followed by early-cycle direct injection of diesel fuel for combustion phasing control at a medium engine load (9-bar IMEP). The technique can also prevent excessive rates of pressure rise.
By varying in real time the fuel reactivity (and therefore, the equivalence ratio) of the charge across the chamber using timed injections of matched fuels—including gasoline and diesel mixtures, ethanol and diesel, as well as blends of gasoline and gasoline with small additions of a cetane-number booster—the investigators have extended the operation of partially PCCI combustion.
“The fuel reactivity is varied from point to point in the combustion chamber,” Reitz said. This is done by injecting diesel into a mixed gasoline/air charge to create the desired gradient. “We can change the effective octane number of the fuel at each location.”
Diesel fuel is more reactive than gasoline and tends to ignite earlier. The relative amounts of the two fuels help determine when combustion starts, how long it takes, and the temperature of burning, he continued. Engine operations would depend on the load: high load—90% gas, 10% diesel; low load—90% diesel, 10% gas; idle—100% diesel.
Genetic optimization points the way
Significantly, the improvements in engine efficiency and emissions performance were realized with the unexpected help and guidance of a computer model. The team's sophisticated software conducts 3-D numerical analysis of the combustion process. The results produced by the CFD model of turbulent flame propagation that includes detailed simulations of chemical kinetics are then optimized by a genetic algorithm. “We run 1000 simulations to select the combustion conditions that give the best operation,” Reitz said.
Starting about two years ago, the team improved its combustion model by adding chemical kinetic rates and advanced spray models. The researchers employ a genetic algorithm that is coupled to the KIVA-CHEMKIN code in which each computational cell is treated as an individual, well-stirred reactor. “This helps us select optimal combustion scenarios, including specific parameters such as injection timings, fuel fractions, and intake valve closing timings” for dual-fuel PCCI combustion, Reitz explained. The group then validates the modeling results with engine experiments.
The veteran engineer acknowledged that no one expected the model to elicit the insights into the combustion process that it has. “We didn’t foresee the efficiency improvements at the beginning."
"Some people are running HCCI, which avoids fuel-rich regions that might produce soot,” Reitz continued. “But you need to keep the temperatures under 1700°K to avoid producing NOx. Others do PCCI, involving various premixed approaches. We’re doing a form of what we call reactivity control charge compression ignition, or RCCI.”
In one test with the truck engine at the 9-bar (130-psi) operating point, NOx and soot were 0.012 g/kW·h and 0.008 g/kW·h, respectively, while achieving 53% net indicated thermal efficiency.
He pointed out that the dual-fuel RCCI approach requires only low injection pressures—500 to 800 bar (7.3 to 11.6 ksi) instead of the 2000 bar (29 ksi) of the common-rail diesel. “That means you can get by at a reduced system cost because a common rail is not needed.” The lab experiments were conducted with conventional common-rail injectors (i.e., wide angle and large nozzle hole).
Gasoline plus additive
The Wisconsin team is also experimenting with a single-cylinder, high-speed automotive diesel from General Motors that burns a different set of fuels. The concept is the same: one fuel—in this case, gasoline—is port-injected and another more reactive one is injected in multiple, precisely timed squirts.
The other fuel, which mimics the effect of the diesel fuel injection, is also gasoline, but it is mixed with small amounts (0.2% of the fueling rate) of a cetane number-improver called di-tert-butyl peroxide (DTBP).
“A smaller, light-duty compression-ignition engine, by definition, has higher heat losses, so we get maybe 5% less—45% to 50%—thermal efficiency,” he noted. The Wisconsin group estimates that an owner of such a dual-fuel car would need to fill a quart-size window washer bottle of the additive during each oil change.
The RCCI process can run with a diesel-ethanol combination as well.
Another focus of future study at the lab will be biofuels. Reitz said that any application of the technique to spark-ignition engines would require some redesign. “Smaller engines tend to have high swirl, but the dual-fuel approach doesn’t benefit from high swirl.”
Multicylinder engine tests
Engineers at Oak Ridge National Laboratory (ORNL) have tested a multicylinder dual-fuel RCCI engine (a modified four-cylinder, 1.9-L European GM diesel), said Robert Wagner, Acting Director of the Fuels, Engines, and Emissions Research Center there. The efficiency improvements are not as large as the single-cylinder power plants, he said, because of the increased complexity of multiple cylinders—turbocharging, a real EGR system, cylinder-to-cylinder and cycle-to-cycle losses.
“The engine needs optimizing, but we think that we understand what we need to do to get to higher efficiencies and low emissions,” he said.
“The important thing is that the new method allows us to run at a lower EGR temperature, which means that the intake temperature can be lower, where the system can handle heat rejection better,” Wagner emphasized. “That leads us to believe that we may be able to push out the load while still keeping the intake temperatures at levels that are compatible with production hardware.”
“Besides increasing your operating range, the extra degree of freedom provided by two fuels seems to make the process fairly robust,” he added. The transition to a multicylinder setting is usually a lot harder, the ONRL researcher noted. “We’re surprised how robust it is.”
The immediate plan is to model the engine at a higher load point and do more controls work to minimize losses as well as hydrocarbons and carbon monoxide emissions. The hope is to run the power plant at 9 bar with 45% peak brake thermal efficiency by end of the year, Wagner said, with large loads afterward. “Higher load capabilities are becoming increasingly important for downsizing and downspecing. Plus, there are ways to do hybrids where you might want to run at higher loads.”
The Engine Research Center’s dual-fuel work is funded by the U.S. Department of Energy and the Diesel Engine Research Consortium, which counts more than 20 industrial members including Caterpillar, Cummins, Ford, GM, Navistar, PACCAR, Renault, TARDEC (U.S. Army), Toyota, and Volvo.
In response to everyone’s first question about the need to replenish two fuel tanks, Reitz said: “It’s not a big issue, fortunately. Trucks already use two fuels—diesel and the urea diesel exhaust fluid—so operators have already learned to put up with that inconvenience.”