By John DeGaspari, Associate Editor

Although diesels are very efficient, the black smoke and particulates in their exhaust have generally kept these engines off the list of clean technologies. One way to reduce particulate emissions is to increase the oxygen content in the combustion air so the engine will burn the fuel more thoroughly. This solution, however, has always had a catch: Reduction in particulates is accompanied by a rise in nitrogen oxide emissions, because of higher combustion temperatures.

Oxygen-enriched combustion experiments were done on a two-stroke, two-cylinder locomotive research diesel engine, in which the oxygen was supplied by an external source.

But now researchers at Argonne National Laboratory in Argonne, Ill., have run tests in which they reduced particulate emissions and NOx at the same time, and increased engine performance. To achieve those results, they followed an operating strategy that included modestly increasing oxygen to between 23 and 25 percent of the engine's air supply, slowing the fuel injection timing, and increasing the rate of fuel flow to the engine.

With the best mix of oxygen, fueling rate, and timing, Argonne achieved a 60 percent reduction in particulates and a 15 percent reduction in NOx, while increasing gross engine power by 18 percent. These results were obtained on an EMD 567B locomotive research diesel engine. In this case, oxygen was supplied from an external source. The lab is currently developing a membrane that would be capable of varying the air composition to supply added oxygen, or perhaps nitrogen, to the engine as needed.

At this point, Argonne is still at a very early stage in the testing of its concept, according to Charles Horton, manager of advanced technologies at General Motors Electro-Motive Division in LaGrange, Ill., an industrial partner in the project. The test engine is a locomotive-size two-cycle, two-cylinder diesel engine with a full-size combustion chamber. Tests are being conducted on a single-cylinder basis, treating parameters as independent variables, said Horton. The tests have not been scaled up to a full-size 12- or 16-cylinder locomotive engine.

Horton regards the tests as promising. "Rather than having a classic tradeoff to reduce NOx and buildup of particulate matter and vice versa, they actually found an operating range where they got both reductions. That's quite a feat and most encouraging."

Michael Rush, associate general counsel for the Association of American Railroads in Washington, Argonne's other development partner for locomotive engines, cautioned that the results are still tentative. He pointed out that the added oxygen came from a separate source, and researchers had reduced emissions on a small test engine in a laboratory. That's a long way from achieving the same results in a 16-cylinder locomotive diesel equipped with an oxygen-enhancing device that is still under development, in his view.

The membrane consists of bundled hollow fibers.

Yet Ramesh Poola, a research scientist with the Energy Systems Division in Argonne's Center for Transportation Research, suggests two reasons why locomotive engines are an excellent test application for the membrane technol-ogy. One is cost: Membranes would represent only a small percentage of the overall cost of big-ticket locomotive engines. Secondly, to meet new emission standards, the U.S. Environmental Protection Agency included locomotive diesel engines in emission reduction goals proposed last year.

The membrane "chemical filter" separates air into oxygen-rich and nitrogen-rich streams.

Current levels of NOx produced by locomotive diesels average 13.5 grams/brake-horsepower-hour. The EPA's goals are to reduce these levels to 7.4 g/bhp-h by 2004 and 5.5 g/bhp-h the following year and after. Particulate levels, which are now at 0.34 g/bhp-h, are below the EPA's goals until 2005, when the goal will be 0.2 g/bhp-h. These goals apply to new locomotive engines as well as rebuilds of diesels originally produced between 1973 and 1999.

Developing a compact on-board means of supplying the oxygen-enriched air to the engine is a key challenge in moving the test results out of the lab and into the real world. This is the focus of a related development effort at Argonne, which is working with Compact Membrane Systems Inc. of Wilmington, Del., to develop a membrane to separate ambient air into oxygen-rich and nitrogen-rich streams before entering the engine.

Prototypes of the membrane exist for passenger car diesels, but they are still being developed for truck and locomotive engines. Poola, Argonne National Laboratory's point man for the development of the membrane, believes that an on-board membrane is the only practical way of introducing oxygen-rich air to a locomotive diesel to adjust its emissions.

Such membranes are not new, and are widely used in the industrial sector for supplying nitrogen in applications such as chemical blanketing, said Poola. However, the availability of new polymers is opening the way for development of membranes for engine applications, he said.

Over the last 10 years, improvements in electronically controlled fuel injection equipment to optimize operating conditions, such as injection pressure and injection timing, have cleaned up diesel exhaust consider-ably, Poola said. The missing element, in his view, has been variable air composition.

"People don't realize that air composition can be varied," Poola said. "They try to optimize the engine configuration, the fueling conditions, or the fuel properties. But the air is also a variable, because the composition can be varied by the membrane." By combining all these factors, greater benefits are achievable, he claimed.

Poola likens his concept of a membrane to a sort of chemical turbocharger, which, instead of pushing more air through an engine, sends more oxygen. This allows the engine to burn more fuel per stroke and produce more power. On the other hand, prototype membranes currently draw from 5 percent to 10 percent of engine power to operate. The power draw results from a certain pressure that must be maintained across the module. The goal is to reduce the power draw to about 2 percent of engine power, he said. Even so, he added, use of such a membrane offers an opportunity to increase power by a minimum of 10 percent.

The oxygen boost delivered to the engine winds up only slightly higher than the 21 percent oxygen content in ambient air, noted Poola. Yet the modest increase in oxygen, together with optimal fueling rate and timing, was enough to make the difference in emissions reductions and increased raw engine power. "Engine combustion is a complex phenomenon," he said. "It needs more oxygen, but it also needs more mixing with the fuel in the combustion chamber."

Membrane technology has come a long way in the last decade in terms of size and power requirements, Poola said. The current version of the membrane measures 7x3 inches, one-tenth the length and drawing 60 percent less power than its predecessors, while producing the same flow and purity, according to Argonne. The lab's goal is to produce a membrane about the size of a car's air filter.

The membrane operates as a solution-diffusion mechanism. Air molecules dissolve into the membrane and then diffuse across it. Air is fed into one side of the membrane at elevated pressure, while the other side of the membrane is kept at lower pressure. The membrane contains a bundle of hollow fibers. Inside each fiber is a thin layer of polymer that selectively absorbs oxygen molecules, which then diffuse out of the other side of the strip. Oxygen-rich air exits the filter on the low-pressure side, while nitrogen-rich air is swept out without crossing the membrane. The ratio of oxygen to nitrogen depends on the polymer being used in the fiber and the pressure differential across the membrane, Poola said.

New developments in materials have opened the way for better membranes. Polymers have traditionally been either rubbery or glassy, offering enhanced permeability or selectivity, but not both. New material families, particularly polyimides, are offering a balance of properties, helping to diminish this tradeoff. CMS is developing a polymer in the polyimide family for engine applications.

Argonne National Laboratory's optimum combination of engine operating conditions reduces the formation of particulates and NOx, and improves fuel economy and gross power.

The membrane should not be adversely affected by water or humidity, according to Poola. Although high temperatures pose a potential seal degradation problem, new seals being tested on prototype membranes can withstand temperatures of 175°C, higher than typical diesel engine service, he said. Higher temperatures, incidentally, improve air separation, he added.

The membrane can degrade from oil or dust particles. "You have to make sure that there are no lubricating oils in the air," Poola said.

Theoretically, a membrane can last for the life of the engine, although under real-world conditions, the device might have to be replaced after a certain period of service. "At this point, we really don't know," he said.

Oxygen enrichment is not the only possibility of air composition control for diesel engines. Here are several other potential areas of development at Argonne for diesel engines.

Nitrogen enrichment: Many light-duty diesel engine manufacturers and truck engine manufacturers attempt to reduce NOx by recirculating exhaust gas back to the intake manifold, which reduces combustion temperatures. Yet exhaust-gas recirculation has downsides, said Poola. Sulfur content and carbon particles in the exhaust gas can contaminate oil and be a source of corrosion and wear. Nitrogen-rich air from a membrane potentially could do the same job as exhaust-gas recirculation, but much more cleanly, he said. Poola is currently testing the nitrogen enrichment concept on a Volkswagen TDI 1.9-liter passenger car diesel engine. Passenger car diesels presently emit an average of about 1.0 grams per mile NOx and 0.08 g/mile particulate matter. The EPA and Partnership for a New Generation of Vehicles, the research arm of the United States Council for Automotive Research in Southfield, Mich., have established targets of 0.2 g/mile of NOx emissions and 0.01 g/mile particulate matter emissions by 2004; the California Air Resources Board has established an even lower number of NOx emissions: 0.05 g/mile (LEV-2).

Variable air composition: The membrane, which produces two streams—one oxygen-rich, the other nitrogen-rich—can be tailored to supply one or the other as the need arises, Poola said. Air composition could be monitored with an oxygen sensor, and be changed on the fly. Under certain load conditions—for example, when a truck pulls away from a signal light—oxygen-rich air can be introduced to reduce black smoke. Under heavy load conditions, when NOx emissions are high, nitrogen-rich air can be introduced. This concept may be suitable for light-duty diesel engines in trucks, said Poola. Heavy-duty diesel trucks must reduce NOx emissions from 4.0 g/bhp-h to 2.0 g/bhp-h by 2004, according to EPA standards.

Late-cycle oxygen-enriched air injection: Within a single engine stroke, soot and NOx form under different conditions during combustion. One air management technique introduces nitrogen- enriched air during the first half of the combustion time, which is prone to produce NOx, and supplies oxygen-rich air during the second half of the combustion cycle, when most particulates are continuously being formed and burned. In this way, nitrogen and oxygen are introduced alternately to suppress NOx and particulates when they are being produced during combustion. Argonne has been investigating this technique for about a year, in partnership with a diesel engine manufacturer and the University of Wisconsin at Madison.


Return to Index