5
Compression-Ignition Diesel Engines

INTRODUCTION

Light-duty compression-ignition (CI) engines operating on diesel fuels have the highest thermodynamic cycle efficiency of all light-duty engine types. The CI diesel thermodynamic cycle efficiency advantage over the more common SI gasoline engine stems from three major factors: the CI’s use of lean mixtures, its lack of throttling of the intake charge, and its higher compression ratios. In a CI diesel engine-equipped vehicle, there is an additional benefit of reduced volumetric fuel consumption (e.g., gal/100 miles) because diesel fuel provides more energy per gallon than gasoline, as is discussed later in this chapter.

Lean mixtures, whose expansions are thermodynamically more efficient because of their higher ratio of specific heats, are enabled by the CI diesel combustion process. In this process, diesel fuel, which has chemical and physical properties such that it self-ignites readily, is injected into the cylinder late in the compression stroke. Ignition occurs following atomization of the fuel jet into small droplets that vaporize and mix, creating pockets of heterogeneous combustible mixtures. These heterogeneous mixtures burn with localized diffusion flames even though the overall fuel-to-air ratio may be too lean to support turbulent flame propagation such as occurs in an SI gasoline engine. This ability to successfully burn overall lean mixtures allows CI diesel engine power output to be controlled through limiting the amount of fuel injected without resorting to throttling the amount of air inducted. This attribute leads to the second major factor enabling the higher efficiency of CI diesel engines, namely the absence of throttling during the intake process, which otherwise leads to negative pumping work. SI gasoline engines must be throttled to control their power output while still keeping the fuel-air ratio at the stoichiometric ratio necessary for proper functioning of their three-way exhaust catalyst. Finally, the diesel combustion process needs higher compression ratios to ensure ignition of the heterogeneous mixture without a spark. The higher CI diesel compression ratios (e.g., 16-18 versus 9-11 for SI gasoline) improve thermodynamic expansion efficiency, although some of the theoretical gain is lost due to increased ring-to-bore wall friction from the associated higher cylinder pressures.

Fuel economy technologies considered in the NRC’s (2002) earlier report on fuel economy did not include diesel-powered CI engines because the costs and emission control systems to meet upcoming nitrogen oxides (NOx) and particulate emission standards were not developed at that time. The motivation for including light-duty CI engine technology in this report stems from two factors. Light-duty CI engine vehicles are now in widespread use in Europe because a high fuel tax on diesel and gasoline fuel allowed diesel retail prices to be substantially lower than gasoline prices. This differential is disappearing in some countries but still persists in others. European buyers have accepted initial higher CI vehicle purchase prices in return for their lower fuel consumption as well as excellent performance and driving dynamics resulting from their high torque. CI diesel vehicles constitute around 50 percent of the new light-duty vehicle market in Europe (DieselNet, 2008). However, in the 2007 U.S. light-duty market, CI diesel vehicles accounted for only about 1.7 percent of the new light-duty vehicles sales (EIA, 2009a). Recent demonstrations of diesel combustion and exhaust aftertreatment systems have shown the capability to meet U.S. 2010 Tier 2, Bin 5 and LEV II emissions regulations for light-duty vehicles. As a result of the emissions control capability achieved by original equipment manufacturers (OEMs) with their internal development projects, at the 2008 Detroit auto show 12 vehicle manufacturers announced the introduction of 13 new CI diesel powered vehicles for the 50-state 2009 U.S. market (Diesel Forum, 2008). However, due to the large fuel price increases of early 2008 and the resulting reduction in vehicle sales of larger vehicles, many OEMs canceled CI vehicle introductions announced for 2009. Nonetheless, four OEMs have offered 12 2009 CI vehicle models.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 61
5 Compression-Ignition Diesel Engines INTRODUCTION ratios (e.g., 16-18 versus 9-11 for SI gasoline) improve thermodynamic expansion efficiency, although some of the Light-duty compression-ignition (CI) engines operating theoretical gain is lost due to increased ring-to-bore wall on diesel fuels have the highest thermodynamic cycle effi- friction from the associated higher cylinder pressures. ciency of all light-duty engine types. The CI diesel thermo- Fuel economy technologies considered in the NRC’s dynamic cycle efficiency advantage over the more common (2002) earlier report on fuel economy did not include SI gasoline engine stems from three major factors: the CI’s diesel-powered CI engines because the costs and emission use of lean mixtures, its lack of throttling of the intake control systems to meet upcoming nitrogen oxides (NO x) charge, and its higher compression ratios. In a CI diesel and particulate emission standards were not developed at engine-equipped vehicle, there is an additional benefit of that time. The motivation for including light-duty CI engine reduced volumetric fuel consumption (e.g., gal/100 miles) technology in this report stems from two factors. Light-duty because diesel fuel provides more energy per gallon than CI engine vehicles are now in widespread use in Europe gasoline, as is discussed later in this chapter. because a high fuel tax on diesel and gasoline fuel allowed Lean mixtures, whose expansions are thermodynami- diesel retail prices to be substantially lower than gasoline cally more efficient because of their higher ratio of specific prices. This differential is disappearing in some countries heats, are enabled by the CI diesel combustion process. In but still persists in others. European buyers have accepted this process, diesel fuel, which has chemical and physical initial higher CI vehicle purchase prices in return for their properties such that it self-ignites readily, is injected into lower fuel consumption as well as excellent performance the cylinder late in the compression stroke. Ignition occurs and driving dynamics resulting from their high torque. CI following atomization of the fuel jet into small droplets that diesel vehicles constitute around 50 percent of the new vaporize and mix, creating pockets of heterogeneous com- light-duty vehicle market in Europe (DieselNet, 2008). bustible mixtures. These heterogeneous mixtures burn with However, in the 2007 U.S. light-duty market, CI diesel ve- localized diffusion flames even though the overall fuel-to-air hicles accounted for only about 1.7 percent of the new light- ratio may be too lean to support turbulent flame propagation duty vehicles sales (EIA, 2009a). Recent demonstrations such as occurs in an SI gasoline engine. This ability to suc- of diesel combustion and exhaust aftertreatment systems cessfully burn overall lean mixtures allows CI diesel engine have shown the capability to meet U.S. 2010 Tier 2, Bin 5 power output to be controlled through limiting the amount and LEV II emissions regulations for light-duty vehicles. of fuel injected without resorting to throttling the amount of As a result of the emissions control capability achieved by air inducted. This attribute leads to the second major factor original equipment manufacturers (OEMs) with their in- enabling the higher efficiency of CI diesel engines, namely ternal development projects, at the 2008 Detroit auto show the absence of throttling during the intake process, which 12 vehicle manufacturers announced the introduction of 13 otherwise leads to negative pumping work. SI gasoline en- new CI diesel powered vehicles for the 50-state 2009 U.S. gines must be throttled to control their power output while market (Diesel Forum, 2008). However, due to the large fuel still keeping the fuel-air ratio at the stoichiometric ratio price increases of early 2008 and the resulting reduction necessary for proper functioning of their three-way exhaust in vehicle sales of larger vehicles, many OEMs canceled CI catalyst. Finally, the diesel combustion process needs higher vehicle introductions announced for 2009. Nonetheless, four compression ratios to ensure ignition of the heterogeneous OEMs have offered 12 2009 CI vehicle models. mixture without a spark. The higher CI diesel compression 61

OCR for page 61
62 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES TECHNOLOGIES AFFECTING FUEL CONSUMPTION two-stage turbocharging (Figure 5.1). Increased boosting is also used for downsizing SI engines. The fuel consumption of engine systems is driven by two Most current light-duty CI diesel engines use a single- major elements, the base engine (i.e., combustion subsystem, stage, variable-geometry turbocharger (VGT). Two-stage friction, accessories, etc.) and the exhaust aftertreatment sub- turbocharger (turbo) systems are being actively developed system. As a result, the fuel consumption of an engine system for two reasons. First, they are a key enabler for engine depends on both the base engine and the aftertreatment. downsizing. Second, they enable increased exhaust gas re- Technologies affecting engine system fuel consumption circulation (EGR) rates. Cooled EGR is the principal method through changes to the base engine and to the aftertreatment to reduce engine-out NOx emissions, as discussed later. With system are discussed below. a two-stage turbo system, two separate turbos are combined with additional flow-control valves. The first-stage turbo is Base Engine Fuel Efficiency Technologies usually sized smaller than the normal single-stage VGT used currently, and the second-stage turbo is usually sized larger The strategies being pursued to improve base engine effi- than the current single-stage VGT. Electronic flow control ciency are the following: valves triggered by the engine controller are used to direct exhaust flows to the small turbo and/or to the large one. At • Downsizing the engine while maintaining equal power, lower engine speeds only the smaller turbo is used and a • Improving thermodynamic cycle efficiency (e.g., im- relatively high inlet pressure is generated, even for the low proved combustion), inlet air flow characteristic of operation at high EGR rates. • Reducing engine friction (e.g., reduced piston skirt At higher engine speeds, when the air flow rates have friction), and increased and the smaller turbo does not have sufficient flow • Reducing accessory loads (e.g., electric water pump, capacity, air flow rates are sufficient to generate high intake reduced fuel pump loads by avoiding fuel recirculation, pressures when the exhaust flow is directed through the larger modulated oil pump). turbo. Therefore, with the use of a two-stage turbo system, the problem of insufficient inlet boost pressure at low speeds Note that all these strategies apply as well to SI engines, with high EGR flow rates is solved without losing engine although the gains may have different magnitudes due to power at high speeds. The ability of two-stage turbo systems process differences between CI and SI engines. to generate higher boost pressures at low engine speeds is the key characteristic of two-stage systems that makes them Downsizing the Engine enablers for engine downsizing. By providing higher intake boost, two-stage systems provide more air in the cylinder, The most significant of these strategies is engine down- thus allowing increased BMEP and torque to compensate sizing, which consists of using a smaller displacement engine for the smaller engine displacement. Naturally, two-stage for a given vehicle mass while still maintaining the same turbo systems are more expensive than single-stage systems. power to give equal vehicle performance.1 This approach To utilize the increased charge mass in the cylinder result- requires higher cylinder pressures (i.e., higher engine brake ing from the higher boost, more fuel must be injected per mean effective pressure [BMEP], which is equivalent to unit of engine displacement. The resulting increased power torque) at any given point on the vehicle drive cycle, which output per unit of engine displacement then compensates reduces engine brake specific fuel consumption (BSFC). To for the downsized engine displacement. Increasing the fuel downsize an engine while still maintaining the same vehicle flow is generally accomplished by increasing the maximum performance, the torque and hence BMEP of the downsized injection pressure, which enables higher injection-pressures engine must be raised at all speeds including the maximum- at all loads. To support the increased cylinder pressures, the power speed. One of the key enablers to raising the BMEP engine structure, sealing (e.g., head gasket), and lubrication is increasing the intake boost provided by the turbocharger (e.g., connecting rod bearings must support higher cylinder system. The emerging approach to increase intake boost is pressures with the same bearing areas) must be improved. Cylinder pressures also increase piston/ring friction, and an additional challenge is to keep the increase to a minimum. 1 Truly equal performance involves nearly equal values for a large number These changes require careful engineering but increase en- of measures such as acceleration (e.g., 0-60 mph, 30-45 mph, 40-70 mph, etc.), launch (e.g., 0-30 mph), gradability (steepness of slopes that can be gine cost only slightly. climbed without transmission downshifting), maximum towing capability, and others. In the usage herein, equal performance means 0-60 mph times within 5 percent. This measure was chosen because it is generally available Improving Thermodynamic-Cycle Efficiency by Optimizing for all vehicles. The equal-performance constraint is important because Combustion and Emissions for Maximum Efficiency vehicle FC can always be reduced by lowering vehicle performance. Thus objective comparisons of the cost-effectiveness of different technologies The combustion process and its phasing relative to piston for reducing FC can be made only when vehicle performance remains motion are important determinants of thermodynamic-cycle equivalent.

OCR for page 61
63 COMPRESSION-IGNITION DIESEL ENGINES FIGURE 5.1 Schematic of two-stage turbocharger system.Figure 5-1.eps low pressure. SOURCE: Joergl et al. (2008). Reprinted HP, high pressure; LP, with permission from SAE Paper 2008-01-0071, Copyright 2008 SAE International. low-resolution bitmap efficiency. However, the combustion process also plays the pends on both cylinder block design and associated thermal key role in the engine-out emissions. As a result, optimizing distortions as well as bore distortion due to mechanical combustion to minimize FC and emissions simultaneously loading by the preloaded cylinder head attachment bolts. requires careful analysis of the interactions between fuel Rounder bores under hot and loaded conditions allow lower spray dynamics, in-cylinder fluid motions resulting from the ring tension, which in turn decrease ring-to-wall friction. interactions of the intake flow with the piston bowl shape Coatings to reduce ring friction are also being developed, (i.e., combustion chamber), gas temperature history, and although it is not yet clear whether such coatings can be chemical reactions of the fuel. As fuel composition evolves both friction reducing and sufficiently durable. Piston skirt from entirely petroleum based to a mixture of petroleum friction can be reduced by improved skirt surface coatings. and bio-sourced components in the next decade to reduce Most current pistons have proprietary skirt coatings, but new petroleum dependence and increase sustainability, it is materials are continuously being studied to further reduce critical that understanding of combustion be increased. It skirt-to-wall friction. is believed that advanced combustion research with tools such as three-dimensional computational fluid dynamic Reducing Accessory Loads computer codes, including spray and combustion as well as coordinated experiments in highly instrumented engines Engine loads to drive accessories include those for cool- with optical access for advanced laser-based tools, will ant pump, oil pump, alternator, air-conditioning compressor, improve understanding of combustion in the longer term. power-steering pump, etc. Electric-motor-driven coolant This improved understanding is critical to reducing exhaust pumps are being considered because they can be turned off emissions without compromising engine efficiency and or run slowly during engine warm-up and at other conditions along with new technologies discussed later should enable when coolant flow can be reduced without engine damage, reductions in FC. thus reducing fuel use to drive the electrical alternator. Two- mode mechanical water pumps are also being developed that require less power to drive at part-load engine conditions but Reducing Engine Friction still provide more coolant flow at high-load conditions. Oil Friction sources in engines are journal bearing friction, pumps, like coolant pumps, are sized for maximum engine valve-train friction, and piston assembly friction. In the past power conditions and are hence oversized for part-load, low- 10 to 15 years, all significant sliding interfaces in valve trains speed conditions. Two-mode oil pumps are being developed have been replaced by rolling interfaces, which minimize and becoming available. friction. Connecting rod, camshaft, and main bearing friction is hydrodynamic, thus coming primarily from lubricating oil Exhaust Emissions Control of CI Diesel Engines shear processes. This friction has been reduced by the use of lower viscosity lubricants. Therefore, the largest remain- The most critical aspect of increasing the use of CI ing friction sources in both CI and SI engines is that due diesel engines in the United States to take advantage of their to the piston assembly. Friction from this assembly comes excellent efficiency is the development and production of from both piston skirt-to-wall interactions as well as piston technologies that can enable these engines to meet the 2010 ring-to-wall interactions. Both skirt and ring friction can be and post-2010 exhaust emissions standards. As noted above, decreased by improved cylinder-bore roundness, which de- CI diesel engines without emission controls have very low

OCR for page 61
64 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES FC characteristics. So the challenge for CI engines is to Siebers, 2004), and others. All these partially homogeneous reduce emissions into compliance without losing the excel- charge strategies drive the combustion process in the direc- lent fundamental CI low FC. This challenge is in contrast to tion of HCCI (homogeneous-charge compression ignition) the case of the SI gasoline engines, for which reducing FC (Ryan and Callahan, 1996). The term HCCI in its purest is the major issue. As noted earlier, in the 2009 model year form refers to virtually homogeneous rather than partially 13 new CI diesel vehicles were announced for introduction homogeneous charge. to the U.S. market (Diesel Forum, 2008). These vehicles To utilize these premixed forms of combustion, a number have been developed to meet the 2010 emissions standards, of measures are used to reduce temperatures and improve and so whatever efficiency deterioration has occurred as a mixing of the charge. The simplest and most effective result of applying the combustion and exhaust aftertreatment measure is increased EGR, as noted above. In addition to technologies necessary to meet the standards is reflected by increased EGR, lowering compression ratio also reduces the fuel economy of these vehicles. Data from the 2009 VW mixture temperatures and, as a bonus, allows increasing Jetta indicated that the fuel consumption reduction between engine power without exceeding cylinder-pressure design the diesel and gasoline versions of the Jetta expected from limits. Lower compression ratios make developing accept- earlier (e.g., 2006) models has been retained, in spite of the able cold-start performance more challenging in spite of significantly reduced emissions, although this result may not improved glow plugs and glow plug controls. hold true for all the new diesel models. As a result, the overall Technologies being developed to support this move in choice between investing in SI gasoline engine technologies combustion technology toward premixed low-temperature to reduce the SI gasoline fleet FC on the one hand and replac- combustion are cylinder-pressure-based closed-loop control; ing some SI gasoline engines with CI diesel engines on the piezo-actuated higher-pressure fuel injectors; two-stage other hand will rest on the total cost for emissions-compliant turbocharger systems; and combinations of high- and low- CI diesel engines and their remaining FC advantage after pressure EGR systems. emissions control measures are implemented. In addition to the specific FC tradeoffs between SI and CI FC, business de- Cylinder-Pressure-Based Closed-Loop Combustion Control cisions on whether to tool up CI engines also depend heavily Technologies on the availability of investment capital in an industry under- going drastic financial problems as well as expectations of Cylinder-pressure-based closed-loop combustion control the willingness of buyers to invest in CI engines, with which technologies enable operating the engine closer to the low- they are largely unfamiliar or have out-of-date perceptions. temperature limit without encountering misfire or excessive hydrocarbon and carbon monoxide (HC/CO) emissions. This technology is especially important in the North Ameri- Combustion System Technologies can market, where the variation of North American diesel The direction for CI diesel combustion system technology fuel ignition quality (i.e., cetane number) is greater than in development has been toward more premixed combustion Europe. This large cetane number variability makes com- and away from traditional CI diesel engine diffusion-type bustion control more difficult especially for more dilute, combustion. Diffusion-type combustion tends to generate lower-temperature combustion strategies. The FC impact both high NOx and high particulate matter (PM) engine-out of cylinder-pressure-based closed-loop combustion control emissions because diffusion flames tend to stabilize at a is 0 to 5 percent. However, since certification fuels are well nearly stoichiometric local mixture ratio that is character- controlled, the efficiency impact would not be observed on ized by high temperatures and resultant high NOx forma- the drive cycle for vehicle emissions certification, but only tion. Surrounding this local stoichiometric diffusion flame in customer use when poor ignition quality fuels are encoun- are rich local fuel mixtures whose thermal and mixture tered in the marketplace. environment also cause high PM formation. Higher levels of dilution by means of large amounts of EGR as well as Piezo-Triggered Common-Rail Fuel Injectors earlier injection and longer ignition delays reduce both average and local temperatures as well as allowing more Piezo-actuated common-rail fuel injectors are being mixing time, thus making the local fuel-air ratios much developed aggressively by the global diesel fuel-injection leaner. This combination of lower temperatures and locally system suppliers (e.g., Bosch, Continental, Delphi, and leaner mixtures minimizes the extent of diffusion flame oc- Denso). These injectors open faster and more repeatably currence and thereby reduces both NOx and PM emissions. than do solenoid-actuated injectors, thereby enabling more The combustion strategies that utilize this approach have injections per combustion event. The latest generations of been given many different names in the literature, including these injectors designed on direct-acting principles entered PCI (premixed compression ignition) (Iwabuchi et al., 1999), low-volume production for the 2009 model year in European PCCI (premixed-charge compression ignition) (Kanda et passenger cars. Multiple injections per combustion cycle al., 2005), LTC (low-temperature combustion) (Pickett and allow lower combustion noise (i.e., diesel knock) and more

OCR for page 61
65 COMPRESSION-IGNITION DIESEL ENGINES precise control of mixing and local temperatures than is pos- This approach was inexpensive and effective in the early sible with a single injection per cycle. This additional level days of CI engine emissions control. However, as emis- of control is useful to maximize the benefits of premixed sion standards tightened, more EGR was needed, resulting low-temperature combustion. In addition to combustion in the hot intake mixture problem noted above. Partly to control, multiple-injection capability is used to enable post- avoid the hot-EGR and EGR cooler fouling problems, low- combustion injections, which have been used as part of the pressure (LP) EGR systems have been developed (Keller engine control strategy used to trigger and sustain regenera- et al., 2008). tion of particulate filters. In low-pressure systems, exhaust gas is taken from the ex- haust system downstream of the particulate filter. As a result, particulates and heavy hydrocarbons have been removed. In EGR Issues addition, these exhaust gases are much cooler since energy Using increased EGR levels to reduce mixture tempera- has been removed by expanding the gases down to atmo- tures to suppress formation of NOx and PM creates two major spheric pressure through the turbocharger turbine and by difficulties in addition to the points mentioned above. First, heat transfer in the exhaust piping leading to the particulate the levels of EGR at idle and part-load conditions typical of filter. As a result, these cooler, cleaner low-pressure exhaust urban and extra-urban driving can reach 60 to 70 percent. gases now have to be pumped back up the intake boost pres- This means that with normal high-pressure EGR, only 30 sure by passing them through the turbocharger compressor to 40 percent of the engine air flow is going through the and subsequently through the charge cooler. EGR systems turbocharger with the remainder recirculated back through combining both high-pressure and low-pressure circuits the engine. As a result, the turbine generates less torque have been developed and put into production on light-duty and the ability of the turbocharger to boost intake pressure vehicles (e.g., the 2009 VW Jetta) (Hadler et al., 2008). is severely hampered. Low inlet pressures lead to lower cylinder charge masses, causing richer mixtures and thus Variable Valve Timing increasing PM formation as well as making it more difficult for post-combustion oxidation of both PM and HC/CO due Some suggestions have been put forth that variable valve to lower oxygen availability. timing (VVT) mechanisms may provide opportunities for The second difficulty associated with very high EGR levels improved usage of EGR as well as other emissions control is that EGR cooling requirements increase. EGR cooling is functionality (Bression et al., 2008) for CI engines. However, extremely important because EGR enters the EGR cooler at the current consensus from advanced development groups exhaust temperatures. Mixing this hot EGR with intake air, at OEMs and consulting firms is that VVT for CI diesels which is already heated through compression in the turbo- provides little or no benefit and therefore is not cost effective. charger compressor, leads to hot inlet mixtures. Hot inlet mixtures negate some of the potential of lowering NOx and Exhaust Aftertreatment Technologies PM formation through lower mixture temperatures. There- fore, high EGR levels require larger and more effective EGR HC/CO Control coolers. Not only do these larger coolers present packaging difficulties in already crowded engine compartments, but they The control of HC/CO has traditionally been relatively also are subject to fouling through condensation of heavy easy for CI engines due to the relatively low levels of these hydrocarbons and water vapor present in the EGR stream, constituents emitted from conventional CI diesel combus- which form deposits inside the EGR cooler decreasing their tion, in spite of relatively low exhaust temperatures. How- cooling efficiency (Styles et al., 2008). ever, that situation has changed as the CI diesel combustion process has been modified to reduce combustion-gas tem- peratures, which reduces exhaust temperatures even further. High- and Low-Pressure EGR Systems As the combustion temperatures have been reduced, HC/CO In most CI diesel engines, EGR is supplied to the in - emissions have risen. The diesel oxidation catalyst (DOC) take manifold directly from the exhaust manifold before was introduced around 1996 to reduce hydrocarbon emis- the turbo. This approach provides high-pressure, high- sions and in turn to reduce the soluble organic fraction of the temperature exhaust gas to the intake manifold. Thus this dilute particulate matter. As a result of the reduced exhaust type of system is called an HP (for high-pressure) system. temperatures noted above, the DOC is being moved closer The HP approach is simple in principle because the exhaust and closer to the turbocharger outlet to increase the tem- manifold pressure is normally slightly higher than the in - perature of the catalyst to increase its conversion efficiency. take manifold pressure. Thus EGR can be passed directly This packaging trend need not significantly increase costs from the exhaust manifold into the intake manifold at a rate but such minimal cost increases are only possible when other controlled by both the EGR flow control valve and the pres- vehicle changes provide the opportunity to modify the engine sure difference between the exhaust and intake manifolds. compartment packaging to allow space for close-coupling

OCR for page 61
66 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES the DOC. In addition, oxidation catalyst coatings are being NOx in the catalyst. As NOx is adsorbed from the exhaust, added to diesel particulate filters (DPFs) and NOx storage adsorption sites on the surface of the coating fill up. Once all catalysts for additional HC/CO control. the coating sites have adsorbed NOx, the NSC is no longer effective at adsorbing additional NOx, which then passes right through the NSC. Therefore, at some point before the Particulate Control catalyst is filled, the NSC must be regenerated to purge Particulate filter control of emissions from CI diesel en- the adsorbed NOx and free the sites to adsorb the next wave gines is presently in use by vehicle manufacturers in Europe of NOx. By supplying the NSC with a rich exhaust stream and the United States. These particulate filters are quite ef- containing CO and hydrogen, the CO and H2 molecules de- fective, filtering out 90 to 99 percent of the particulates from sorb the NOx from the catalyst surface and reduce the NOx the exhaust stream, making CI diesel engines more attractive to N2, H2O, and CO2. Therefore, like the particulate filter, the from an environmental impact point of view. Obviously, NSC operates in a cyclic fashion, first filling with NOx from particulates accumulate in the filters and impose additional the lean diesel exhaust (i.e., an oxidizing atmosphere) and back pressure on the engine’s exhaust system, thus increasing then being purged of NOx in a rich exhaust (i.e., a reducing pumping work done by the engine. This increase in pump- atmosphere) that, with the help of precious metals also part ing work increases fuel consumption. In addition, there is of the catalyst surface coating, reduces the NOx back to N2. a second fuel economy decrement caused by the additional Accordingly, application of an NSC to any engine that fuel required to regenerate the filter by oxidizing retained has a lean exhaust stream like diesel engines requires that particulates. The low exhaust temperatures encountered in periodically (every 30 to 60 seconds depending on the size light-duty automotive applications of these filters are insuf- of the catalyst and the operating condition of the engine) the ficient to passively oxidize the accumulated particulates. As a engine system must create a rich exhaust stream for 10 to 15 result, temperatures must be increased by injecting fuel (most seconds to clear the catalyst surface of NOx, thus preparing frequently in the engine cylinder after combustion is over) it to adsorb the next wave of NOx. One approach to creating to be oxidized, raising the temperature of the cylinder gases. the required rich exhaust stream in the engine cylinder is by These hot gases then pass from the cylinder out into the throttling the engine to reduce airflow, thus enriching the exhaust system and then downstream to the particulate filter mixture in the cylinder. Although gasoline engines operate to oxidize the particulates retained in the filter. To achieve quite happily with rich mixtures, operating a CI diesel engine sufficiently rapid regeneration for practical use in light-duty with a rich mixture without forming excessive particulate and vehicles (e.g., in around 10 to 15 minutes), exhaust gases hydrocarbon emissions is quite challenging. If the combus- must be raised to 625 to 675°C. tion process is carried out at sufficiently low temperatures, Engine control algorithms for filter regeneration not particulate formation is minimized, but both hydrocarbon only must sense when the filters need to be regenerated and emissions and FC increase significantly during this brief bring about the regeneration without overheating the filter, rich operation. but also these algorithms must contend with other events An additional difficulty with NSCs is that the catalyst like the driver turning off the vehicle while regeneration is coatings preferentially adsorb sulfur compounds from the underway, thus leaving an incompletely regenerated filter. exhaust. These sulfur compounds originate mostly from When the vehicle is then restarted, the control algorithms the sulfur in the fuel. This sulfur takes up the adsorbing sur- must appropriately manage either completion of the regen- face sites on the catalyst, leaving no sites to adsorb NOx. This eration or start of a new filling and regeneration cycle. These sulfur adsorption, termed sulfur poisoning, is problematic algorithms have become quite sophisticated, with the result even with today’s low-sulfur (<15 ppm) diesel fuel. Some of that particulate filter systems are quite reliable and durable. the sulfur in the exhaust gases may also come from the en- gine lubricating oil. Thus the NSC must also be periodically regenerated to clear out the adsorbed sulfur. Sulfur forms a NOx Control much stronger bond with the catalyst surface than does NOx There are two approaches to aftertreatment of NOx emis- and as a result, sulfur regeneration requires not only a rich sions: NOx storage and reduction catalysts (NSC), which are exhaust stream but also higher temperatures like ~650°C also called lean NOx traps (LNT) (Myoshi et al., 1995), and rather than the typical 200 to 300°C temperatures adequate selective catalytic reduction devices. for NOx regeneration. While the sulfur regeneration does not need to be done nearly as frequently as NOx regeneration, sulfur regeneration also causes a FC penalty. NOx Storage Catalysts The current NOx aged conversion capability of NSCs NOx storage catalysts utilize a typical monolith substrate is around 70 percent. Early attempts to develop NSCs had that has both barium and/or potassium as well as precious difficulty achieving even 50 percent aged conversion ef- metal (e.g., platinum) coatings. These coatings adsorb NOx ficiency in spite of ~80 percent for a fresh NSC. Extensive from the exhaust gas stream to form nitrates, thus storing the development on catalyst test benches indicated that exces-

OCR for page 61
67 COMPRESSION-IGNITION DIESEL ENGINES sive temperatures, particularly during sulfur regeneration, 93 percent or more without the increased engine-out hydro- caused the observed deterioration in conversion efficiency. carbon emissions and FC resulting from NSC regenerations. Recently, two factors have enabled improvements. First, As a result, vehicles using SCR have better FC characteristics newer catalyst formulations have been developed to allow at equivalent emission levels than those using NSC systems. sulfur regeneration at somewhat lower temperatures. Second, When urea is used to provide the ammonia, the urea-water empirical models of catalyst behavior have been developed mixture that is injected into the exhaust stream must be car- and incorporated into the engine controller. The combined ried on board the vehicle. The amount of urea that needs to effect of these two developments has enabled increasing aged be supplied to the SCR catalyst depends on the level of NOx conversion efficiency to ~70 percent. In the summer of 2008, in the exhaust and therefore depends on driving conditions, VW released the 2009 Jetta TDI for the U.S. market which but for light-duty vehicles it is a small fraction of the fuel utilizes an NSC and meets Tier 2, Bin 5, as well as LEV II flow. Initial discussions regarding the possibility of using an emissions standards, enabling VW to sell the vehicle in all 50 SCR-urea approach to NOx aftertreatment for the U.S. mar- states and Canada. A schematic of the aftertreatment system ket were met with concern on the part of the EPA that there used on this vehicle is shown in Figure 5.2. was considerable risk that drivers would not keep their urea tanks filled thus rendering the system ineffective. However, together with EPA oversight, vehicle manufacturers have Selective Catalytic Reduction developed systems to monitor the supply of urea in the urea Selective catalytic reduction (SCR) was originally devel- tank, which will not allow the engine to restart more than a oped for stationary power plants but is now being applied to small number of times (e.g., 20) when the urea supply starts heavy-duty truck CI engines in Europe (Müller et al., 2003) running out, following appropriate warnings to the driver. As and in the United States in 2010. SCR was also introduced a result of such safeguards, the EPA has approved the certi- in the United States in 2009 on some Mercedes, BMW, and fication of the 2009 vehicles using the SCR-urea approach VW vehicles. This system, called BlueTec, was jointly de- to NOx aftertreatment. One example of an SCR-urea-based veloped by all three manufacturers. SCR works by having exhaust aftertreatment system is illustrated in Figure 5.3. ammonia in the exhaust stream in front of a copper-zeolite or iron-zeolite SCR catalyst. The ammonia gets stored on the Combined NSC and SCR Systems catalyst surface where it is available to react with the NOx over the catalyst converting the NOx into N2 and water. To Another strategy that has been proposed is to use a system provide ammonia to the exhaust stream, a liquid urea-water in which the NSC is followed by SCR without external urea mixture is injected into the exhaust sufficiently upstream of addition. It is well known that under some operating condi- the SCR catalyst unit and before a mixer, to allow time for tions with the appropriate washcoat formulation, NSCs can vaporization and mixing of the urea and creation of ammonia convert NOx to ammonia, which is undesirable for an NSC- from the urea, which is an industrial chemical used primarily only system and hence must be cleaned up before exiting the as a fertilizer. In the fertilizer application, urea is relatively exhaust system. However, by following the NSC with SCR inexpensive, but for use with an SCR system, it must be con- without urea injection, which is generally called passive siderably more pure and as a result is more expensive. SCR SCR, SCR will capture and store the ammonia generated by systems tend to have NOx conversion efficiencies of 85 to the NSC and use it to reduce NOx. Since the amount of am- FIGURE 5.2 Exhaust aftertreatment system on the 2009 VW Jetta using NOx storage and reduction catalyst technology for control of NOx. SOURCE: Courtesy of Volkswagen AG. Figure 5-2, fixed image

OCR for page 61
68 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES FIGURE 5.3 Schematic of a BMW exhaust aftertreatment system with selective catalytic reduction (SCR) for NO x control using urea (called AdBlue) addition. The catalyzed soot filter (CSF) is close-coupled to the engine. SOURCE: Mattes et al. (2008). Reprinted with permission. Figure 5-3.eps bitmap FUEL CONSUMPTION REDUCTION POTENTIAL monia generated by the NSC is not large, the passive SCR unit will have low conversion efficiencies but can be a useful CI Fuel Consumption Reduction Advantage supplement to the NSC system. This approach has been used by Mercedes in its Blue-Tec I system used in Europe. In a study for the EPA (EPA, 2008), Ricardo, Inc., car- ried out full system simulation (FSS) to assess the FC and Choosing Between NSC and SCR Systems CO2 impact of many of the technologies expected to enable reduced FC by 2020. FSS calculations were made for the There are both cost and functionality differences between 2007 model-year light-duty vehicle fleet for a set of vehicles NSC and SCR systems which would influence which choice representing five vehicle classes. Combinations of technolo- an OEM might make for NOx aftertreatment with CI engines. gies deemed to be complementary were applied to baseline NSC systems use much more PGM (platinum group metals) vehicles considered to be representative of each class. For than do SCR systems. (The SCR unit itself uses no PGM.) the selected combinations of power train and vehicle tech- As a result, NSC system costs increase faster with increas- nologies, final drive ratios were varied to find the ratios that ing engine displacement than do SCR systems. Thus, from enabled performance equivalent to the baseline vehicles a cost point of view, NSC systems would be chosen for based on a comprehensive set of performance measures smaller displacement engines for which the current 70 per- while minimizing FC. CI diesel power trains were evaluated cent NOx conversion efficiency of the NSC is sufficient to among the combinations of technologies considered. Results reduce engine-out NOx levels to below the Bin 5 emissions for the CI diesel power train CO2 emissions and FC versus standards. As engine displacement is increased and engine- the baseline vehicles for three of the five vehicle classes are out NOx emissions increase, there is an engine displacement summarized in Table 5.1. CI power trains were not applied above which the 70 percent conversion efficiency of NSCs is to the other two vehicle classes, but the results for the three insufficient and the higher (approximately 85 to 93 percent) classes for which CI engines were evaluated are considered conversion efficiency of SCR is required. If PGM commod- representative of all classes. ity prices are sufficiently low, NSC systems costs for larger As indicated in Table 5.1, for the three vehicle classes con- displacement I4 engines (e.g., 2.5 to 2.8 L) might be lower sidered, the average reduction in CO2 emissions was about than those for SCR systems for those same engines, but NOx 23 percent and the corresponding average reduction in FC conversion efficiencies might not be high enough to meet the was 33 percent when the baseline 2007 model year SI power standards. Thus, the engine displacement above which an trains were replaced with CI power trains utilizing DCT6, OEM would choose SCR rather than the NSC is not simply EACC, HEA, and EPS. The 2009 VW Jetta was introduced a cost-based decision. with a 6-speed DSG (VW’s name for DCT6) transmission.

OCR for page 61
69 COMPRESSION-IGNITION DIESEL ENGINES TABLE 5.1 Estimated CO2 and Fuel Consumption Reductions for Three EPA Vehicle Classes, as Determined from Full System Simulation (FSS) Combined SI to CI CO 2 Combined Fuel Combined Fuel Technology Downsize Emissions Consumption Combined Consumption Vehicle Package Major Features Ratio g/mi. gal/100 mi. CO2 Reduction Reduction Full-size Baseline 3.5-L V6 gasoline 356 4.051 Baseline Baseline car SI, AT5 5 2.8-L I4 diesel, 80% 273 2.707 23.3% 33.2% DCT6, EACC, HEA, EPS Small Baseline 2.4-L I4 gasoline SI, 316 3.596 Baseline Baseline MPV DCP, EPS, AT4 5 1.9-L I4 diesel, 79% 247 2.449 21.8% 31.9% DCT6, EACC, HEA, EPS Truck Baseline 5.4-L V8, gasoline 517 5.883 Baseline Baseline SI, CCP, AT4 5 4.8-L V8 diesel, 89% 391 3.877 24.4% 34.1% DCT6, EACC, HEA, EPS Average CI diesel 23.2% 33.0% versus gasoline NOTE: See Chapters 2 and 8 for more information on FSS. To determine the FC reductions, the CO 2 emissions results taken from EPA (2008) were converted to volumetric FC using conversion factors from EPA (2005). AT5, lockup 5-speed automatic transmission; AT4, lockup 4-speed automatic transmission; CCP, coordinated cam phasing; DCP, dual (independent) cam phasing; DCT6, dual-clutch 6-speed automated manual transmission; EACC, electric accessories (water pump, oil pump, fans); EPS, electric power steering; HEA, high-efficiency alternator. SOURCE: Based on EPA (2008). Note also that CI engines were downsized in displacement ductions achievable from engine replacement alone without by an average of about 83 percent from the SI engines they a simultaneous transmission change to DCT6 (and EACC replaced. Tables 7.13, 7.15, and 7.18 from EPA (2008) for with HEA) would be about 25 percent. small MPVs, full-size cars, and trucks, respectively, indicate that these CI engine-powered vehicles with DCT6 transmis- Fuel Volumetric Energy Effect sions provided equivalent performance to the vehicles with larger-displacement original SI engines and transmissions. It should be noted that part of the volumetric FC benefit The 2007 model-year baseline vehicles were equipped of CI diesel engines stems from the differences in volumetric with 4- and 5-speed automatic transmissions. As noted energy content between gasoline and diesel fuels. The energy above, the 33 percent FC reduction indicated in Table 5.1 content of a gallon of diesel fuel is about 11 percent higher reflected DCT6 transmissions and more efficient engine than that of gasoline. While this factor can be an advantage accessories as well as the engine change. To estimate the for drivers if diesel fuel is selling at gasoline prices or lower, separate effect of replacing SI engines and transmissions the carbon dioxide emissions advantage for the diesel would by CI engines with equivalent transmission technology and be less than would be indicated by the volumetric FC advan- without advanced accessories, a European database of 2009 tage of the CI diesel engine. As indicated in Table 5.1, the vehicles was analyzed. Using vehicles that are offered with CO2 reduction advantage for CI engines is about 10 percent 5- and 6-speed transmissions for both SI and CI engines, an less than their FC reduction advantage. estimate was derived of the reduction in FC from replacing SI engines with CI engines at equivalent vehicle performance Fuels for CI Engines without the effect of simultaneously converting from 4- and 5-speed automatics to DCT6 transmissions. The data used for The performance and emissions of diesel engines are this estimate are plotted in Figure 5.4 and shown in tabular also influenced by the fuel characteristics and fuel quality. form in Table 5.A.1 in the annex at the end of this chapter. Although fuel is not a focus of this report, several relevant Figure 5.4 indicates that the average FC reduction for this characteristics for performance and emissions are important vehicle subset was about 25 percent. Therefore, the FC re- in connection with their influence on engine performance,

OCR for page 61
70 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES 40 35 30 % FC Reduction 25 20 15 10 5 0 de 3 0 B M i A3 Au 8 rd sta M d e XF A6 ER o E 0 da y da ic gu V 0 ta is W4 R uge t ta La 8 Au n a e r Au 7 r d g er AV on la x Q A 52 AV an yo 35 0 Ja R- AG on Civ yo ar gu ce E2 au t 3 Je Fo Fie di di c e ar To a Y di Vi C To s S d en R a W o Au G s Av s t de V lt Fo ge ce Pe H H en er er od M M D Vehicle FIGURE 5.4 Percent reduction of fuel consumption (FC) on the NEDC driving cycle for a subset of 2009 European vehicle platforms offered with both SI and CI engines. The subset was selected Figure 5-4.eps from a larger set of 2009 vehicle platforms offered with both SI and CI engines by including only those platforms for which 0-62 mph (0-100 km/mile) times were within 5 percent, which was considered to be equivalent performance. The data used to construct this figure are shown in Table 5.A.1 in the annex at the end of this chapter. enforced and quality is adequate. Cylinder-pressure-based efficiency, and emissions. These characteristics are cetane closed-loop control, as discussed earlier and utilized in one number (a measure of fuel self-ignition in the CI cycle— of the new 2009 CI diesel vehicles, can adjust for market important in cycle efficiency, but also in low-temperature variability in the cetane number of the fuel and provide com- operation), density/heating value (a measure of volumetric pensation over the entire operating engine map. The lower energy content), lubricity (important for fuel system wear lubricity of the U.S. diesel fuel requires protective coatings and durability), and sulfur level (important for proper opera- for the high-pressure pump in the fuel injection system. As tion of the engine exhaust aftertreatment system). noted earlier, the ultralow level of sulfur in the fuel regulated In the U.S. market, there is only one diesel fuel suited for to less than 15 ppm is a necessary enabler for the efficient on-road transportation; its characteristics are specified by and durable operation of the exhaust aftertreatment system. the ASTM Standard D975. Most state regulations require Nonetheless, all OEMs marketing CI diesel vehicles in the the enforcement of these specifications. In the EU, where North American (NA) market have concerns over the sea- light-duty CI diesel passenger cars are widespread and about sonal and regional variability of diesel fuel as well as the half the new cars are powered by diesel engines, the diesel enforcement of fuel quality. fuel is specified by the EN590 standard. There are significant At present, the ASTM D975 fuel standard allows up to 5 differences between the EU and the ASTM standards. The percent biodiesel blend stock in the fuel provided the blend EU fuel has much higher cetane (e.g., 52 versus 40-48), the stock meets the characteristics of the ASTM standard. The fuel density is limited to a minimum to assure adequate en- European OEMs exporting diesel vehicles to the United ergy density (no limit exists in the ASTM standard), and the States have stated that their engines are robust to this fuel lubricity is better. In terms of fuel sulfur, European fuel has blend and that performance and emissions are not affected as similar levels to U.S. fuels, for which sulfur level is regulated long as the blend is at or under 5 percent. For the European by the 2006 EPA standards to 15 ppm or less. market, the manufacturers may allow up to 7 percent FAME In the near future, most diesel passenger cars in the United (fatty acid methyl ester), plus up to an additional 3 percent States will be imports from Europe. Their engines have been hydrogenated biofuel. The difference in the proportion al- adapted for use of U.S. diesel fuel, and the manufacturers lowed by the European OEMs for the U.S. market versus for do not expect to encounter performance and emission issues the European market is due to their concern over the qual- connected with the fuel, as long as fuel specifications are

OCR for page 61
71 COMPRESSION-IGNITION DIESEL ENGINES ity and stability of American blend stock and the variety of is called “downspeeding.” Another 2 to 3 percent is expected feedstocks, including soy, recycled used oils, fats, etc. from reduced transmission internal losses. Efficiency Improvements from Transmissions Overall Fuel Consumption Reduction Potential The transmission technology utilized in the FSS results The FC reduction potential via replacement of SI gasoline shown in Table 5.1 was a dual-clutch 6-speed (automated power trains by base-level CI power trains is illustrated by manual) transmission (DCT), which is a very efficient design Table 5.1 (i.e., ~33 percent) for CI engines with advanced concept. Transmissions used for CI diesels must be designed transmissions (plus EACC, HEA, and EPS) and by Figure 5.4 to handle their larger torque, which may reduce their effi- for engine replacement alone (i.e., ~25 percent). Additional ciencies slightly due to larger gears, bearings, and seals. technical improvements, as noted earlier, from downsizing, DCTs are already in production for smaller displacement CI thermodynamic improvements, friction reduction, and en- engines (e.g., 2009 VW Jetta). The most challenging aspect gine accessory improvements, are being developed and will of designing DCTs with the higher torque capacities needed be implemented. CI engines with these technologies imple- for larger displacement CI engines is providing adequate mented are termed advanced-level CI engines. Transmission cooling for their wet clutches (i.e., oil-cooled clutches). improvements are also possible. Dual-mass flywheels, which reduce drive train vibration, Based on interactions with OEMs, consulting companies, thus reducing heat-generating clutch slippage, will be used. review of the technical literature, and the judgment of the Nonetheless, it is not presently known when such DCT units committee, estimates of the overall FC reduction potential will be available with 500-650 N-m torque capacities for from these advanced-level technology areas are presented in larger CI engines. Table 5.2. For the ranges shown, the 10 percent for engine Expected transmission-based CI vehicle efficiency im- technologies alone and 13 percent for vehicles applies to provements beyond those already comprehended by the use larger vehicles with automatic transmissions. For smaller ve- of the DCT6 transmissions are estimated at 1 to 2 percent for hicles with manual transmissions and engine displacements downspeeding the engine by increasing the number of dis- less than 1.5 L, cost constraints are likely to reduce the extent crete speed ratios beyond six. The increased number of ratios of downsizing and the potential would be about 6 percent for allows keeping the average engine speed lower while still engine alone and 7 percent for vehicle due to elimination maintaining equal performance, which is why this approach of not only the gain from automatic transmission efficiency TABLE 5.2 Estimated Fuel Consumption Reduction Potential for Advanced-Level CI Power Trains Compared to Base- Level CI Power Trains Item Average Reduction (%) Min Max Large Vehicles Downsizing 4 3 5 Downspeeding 1.5 1 2 Friction reduction 1.5 1 2 Combustion improvement 3 2 4 Total engine improvement 10 Accessory improvement 1 0.5 1.5 Transmission loss reduction 2 1.5 2.5 Combined engine and transmission potential 13 Item (%) Reduction Min Max Small Vehicles (<1.5 L) Downsizing 1 0 2 Downspeeding 0.5 0 1 Friction reduction 1.5 1 2 Combustion improvement 3 2 4 Total engine improvement 6 Accessory improvement 1 0.5 1.5 Thermal management 0 0 0 Transmission loss reduction 0 0 0 Combined potential 7 NOTE: The values shown for the combined potential do not show a range. It is tempting to use the sum of the minimum values for the lower limit of the range and the sum of the maximum values for the upper end of the range. However, this would be inappropriate because no original equipment manufacturer is likely to simultaneously achieve either the minimum or the maximum for all items. Therefore, a realistic range for the combined potentials is about ±1 percent.

OCR for page 61
73 COMPRESSION-IGNITION DIESEL ENGINES to realizing some of the efficiency gains summarized in CI vehicles already in the market. These new engines would Table 5.2. For the OEMs active in the European market, this probably be introduced in both base-level and advanced-level timeline is compatible with tax incentives expected in 2011 technology versions in order to both be technologically com- for early introduction of vehicles meeting Euro 6 as well as petitive with advanced-level technology products already in with the next European fleet CO2 reduction target in 2012. the market and to achieve market volumes necessary to jus- The second path for introduction of the advanced-level tify the tooling investment. Third, fuel prices must increase technologies summarized in Table 5.2 is their introduc - from late 2009 levels but without significant negative price tion simultaneously with new CI power trains in the period differential between gasoline and diesel in order to provide 2014-2020. These advanced-level versions will be required potential customers with sufficient incentive to offset the for market competitiveness for these new vehicles since additional prices that must be charged for CI engines. the OEMs introducing CI vehicles between 2009 and 2011 will probably have already implemented advanced-level TECHNOLOGY COST ESTIMATES technology features. For example, BMW has already intro- duced an engine with two-stage turbocharging, one of the There are a number of complexities in making cost esti- key features of the advanced-technology level. However, mations for CI engines to replace SI engines. The first of the pace of introduction of these vehicles with newly tooled these involves selecting the appropriate displacement for the CI engines will follow the new market conditions based on CI engine. This is important because CI engine costs depend the economic recovery of global economies and the related significantly on their displacement for two primary reasons. automobile markets. First, the configuration and cost of their exhaust aftertreat- In addition, California Air Resources Board (CARB) ment systems depend on engine displacement since com- LEV III standards are expected for 2013. The LEV III ponent substrate (e.g., oxidation catalyst, particulate filter) emissions levels currently under discussion would be very volume is proportional to engine displacement and precious challenging. So OEMs will be developing technologies to metal washcoat weights applied to the substrates are propor- enable their diesel products to meet LEV III and associated tional to substrate volume. In addition to washcoat factors, regulations. Studies at European OEMs with development NSC (NOx storage catalyst) and urea-SCR-based NOx reduc- vehicles using emissions control technologies developed to tion systems have different relationship multipliers to engine meet Tier 2, Bin 5 standards indicate that these technologies displacement. This is because urea-SCR-based systems use need additional development to achieve proposed LEV III much less PGM compared to NSC-based systems, thus de- requirements. As a result, it is expected that there will be creasing the rate at which costs increase with displacement. some fuel consumption increase in order to meet the new Second, the degree of downsizing employed for the CI standards. engine determines the cost and complexity of the air system In summary, the following technology sequencing is for the engine. Maximum downsizing corresponding to envisioned: advanced-level CI engines requires two-stage turbo systems, which cost about twice those of base-level single-stage turbo • For OEMs with existing CI engines, vehicles intro- systems. duced in 2009 will be joined by additional models from The cost of the engine structure and mechanical parts 2011 to 2014, with base-level or advanced-level tech- of CI engines depends less on displacement since smaller nology features depending on each OEM’s particular engines have all the same parts as larger displacement ones. marketing strategy. These parts all require the same casting, fabrication, and • During the period 2015-2020, it is expected that de- machining processes and differ primarily in the amount of velopment efforts for these OEMs will be focused on raw materials used, which has a relatively small influence further reduction of power train cost and fuel consump- on total cost. In the present work, no displacement-based tion to achieve the upper limits of the ranges shown in adjustment was made to the cost estimates for the basic Table 5.2. engine structure and parts. For OEMs without existing CI engines with displace- Engine Sizing Methodology ments in the range that would have the biggest impact on improving their CAFE values (e.g., V6 engines with dis- The engine sizing methodology developed for this work is placements around 3.5 L for SUV and pickup trucks), new based on current and future product development directions. engines may be developed and put into production if three Two CI engine configurations have been considered, namely, conditions are met. First, overall light-duty markets in the base-level engines and advanced-level engines, as discussed 2010-2012 period must improve sufficiently from those of above in the subsection titled “Overall Fuel Consumption 2009 to generate improved corporate financial health and re- Reduction Potential.” Performance of a given vehicle de- quired tooling capital. Second, a favorable customer percep- pends primarily on the combined effect of the torque curve tion of CI power trains must evolve based on the 2009-2012 of the engine, the transmission characteristics (e.g., speed

OCR for page 61
74 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES ratio range and internal efficiency), and final drive ratio. For resultant BOM and cost sets in extensive detail. The resultant base-level CI engines, a maximum specific torque density of BOMs included not just the CI engine hardware added or SI 160 N-m/L is assumed. This level is achievable with single- hardware subtracted but also additional components that, in stage turbo systems and, for example, is the level achieved the judgments of the OEMs and suppliers, were necessary by the Tier 2, Bin 5-compliant 2009 VW Jetta. The CI en- to make fully functional vehicles meeting both emissions gines considered in the Ricardo, Inc., FSS analysis (EPA, standards and customer expectations. Martec reviewed the 2008) from which the fuel consumption reduction values resultant cost tables with both the OEMs and the suppliers to in Table 5.1 were determined had base-level technology reach consensus. It is often said by OEMs that cost numbers features with single-stage turbo systems. provided by suppliers are lower than what OEMs actually For advanced-level CI engines, a specific maximum have to pay, while suppliers counter that the costs that OEMs torque density of 200 N-m/L is assumed. This level allows say they have to pay include more content than that quoted downsizing from base-level CI engines, thereby enabling by the supplier. It is hoped, therefore, that the approach additional fuel consumption reductions. The Tier 2 Bin 5 used by Martec to reach consensus avoided this potential compliant 2009 BMW 335d with two-stage turbocharging confusion and provided more correct estimates. Finally, the achieves over 192 N-m/L and the Mercedes OM651 re - Martec study was carried out in 2007-2008—more recently cently introduced in Europe achieves 233 N-m/L, and so the than the years (2002-2006) on which the EPA (2009) es- 200 N-m/L assumed for the advanced-level technology CI timates were based or the period covered (2005-2008) in engine is considered realistic. Duleep (2008/2009) estimates. Based on the results from the full system simulation ve- To avoid the rather subjective issue of cost reductions hicle simulations carried out by Ricardo, Inc., for the EPA over the production life of components, Martec developed (EPA, 2008) (see Table 5.1) for 2007 model-year midsize cost estimates assuming very large production volumes so MPV, full-size car, and truck-class vehicles, base-level CI that all volume-related learning could be considered already engines displacing about 83 percent of the SI engines they reflected by its cost estimates. For some existing compo- replaced achieved equivalent vehicle performance when nents, like common rail injection systems, global produc- combined with advanced DCT6s (6-speed dual-clutch trans- tion volumes are already high enough to exceed the Martec missions). It is therefore assumed that base-level CI engine volume threshold, and cost estimates for these items would displacement is about 83 percent of that of the 2007 model- automatically include cost reductions from high-volume year SI engine being replaced. Similarly, advanced-level CI learning. On the other hand, it is not expected that the CI engines having displacements about 80 percent of those of diesel engines used for the NA market alone will exceed base-level CI engines can maintain equivalent vehicle perfor- that volume threshold before 2020. However, since many of mance. This is because the maximum torque of a base-level these engines will also be produced for the European Union CI engine of displacement d would be about 160 × d N-m. (EU) market, whether by EU OEMs or by U.S. domestic Since the base-level maximum specific torque of 160 N-m/L OEMs that produce such engines for their EU products, the is 80 percent of the 200 N-m/L for the advanced-level CI combined EU, U.S., and Canadian volumes may reach the engine, the appropriately sized advanced-level CI engine 500,000-unit threshold. Thus the volume thresholds required would have 80 percent of the displacement of the base-level to realize high-volume earnings will consist of combined EU engine (i.e., 80 percent × d). Then peak specific torque of and NA volumes for a number of the engines in the CI diesel the advanced-level CI sized at 80 percent would be equal to fleet. It is expected that volumes will reach the 500,000-unit that of the base-level (i.e., 200 × (80 percent × d) ≈ 160 × d). threshold primarily for the engines sold in the highest vol- With equal maximum torque, the advanced-level CI engine umes in the EU (e.g., ~1.6 L). Thus for some of the smaller would enable equivalent vehicle performance. engine displacements likely to have low volumes in the U.S. market (e.g., <1.5 L) as well as for larger engines (e.g., 4.0- 4.5 L) used in vehicles not marketed at high volume in the Cost Estimation Methodology EU (e.g., large SUVs and pickups), the 500,000-unit volume The cost estimations from the sources considered in the target may not be reached by 2020 and costs will remain present work (Martec Group, Inc., 2008; EPA, 2008, 2009; somewhat higher. To that extent, some of the Martec CI cost Duleep, 2008/2009) are then compared with those used by increment estimates could be too low. the NHTSA in its final rulemaking for 2011 (DOT/NHTSA, The cost estimates developed in the present work were 2009). The Martec study used a BOM (bill of materials) derived primarily from the Martec study (Martec Group, Inc., a pproach based on technology packages consisting of 2008). This choice was made for the reasons stated above. combinations of components that fit together technically In addition, the Martec report included detailed specifica- and made sense from a marketing point of view. BOM is tion of the exhaust aftertreatment system configuration, siz- also discussed in Chapter 3. This assessment was made by ing, and PGM washcoat loadings. This type of information OEMs and suppliers with which Martec met. Martec then was not included in EPA (2008, 2009) studies or in Duleep developed component-by-component costs and described the (2008/2009). In addition, the Martec report described the

OCR for page 61
75 COMPRESSION-IGNITION DIESEL ENGINES commodity cost basis used, thus allowing modification of requirements with NSC-based systems and changes in PGM those costs in the present work to reflect recent decreases in commodity prices. However, for the heavier SUV, SCR-urea commodity pricing for PGMs. with its capability for 85 to 93 percent conversion efficiency will be required for emissions compliance. As a result, there are no entries in Table 5.6 for NCS NOx aftertreatment for the Base-Level Engine Technology Cost Estimates SUV since it is assumed that SCR technology will be used. Incremental CI diesel engine cost estimates developed in Commodity prices were quite volatile between 2004 and the present study for replacing 2007 model-year SI gasoline 2008 (Martec Group, Inc., 2008), making product planning engines with equivalent performance CI diesels are summa- for CI diesel vehicles quite challenging. To illustrate the rized in Tables 5.4, 5.5, and 5.6. Appendix G contains the impact of PGM (platinum group metals consisting of plati- same information for full-size body-on-frame pickup trucks. num, palladium, and rhodium) commodity price volatility, Table 5.6 includes estimates for the precious metal wash coats used in the catalysts in separate rows labeled PGM Emissions Systems Cost Estimates loading. In addition, two columns are shown for each of the Since the exhaust emissions systems are a significant frac- two reference vehicles. Columns two and four correspond to tion of the cost for CI diesel power trains, the brief entries in the PGM prices in November 2007 used in the Martec study Tables 5.4 and 5.5 are described in more detail in Table 5.6. (Martec Group, Inc., 2008). The estimates in columns three Note that the entries in Tables 5.4 and 5.5 reflect choices and five illustrate emissions systems costs based on PGM made for NOx aftertreatment technologies. For the midsize prices from April 2009 computed in the present study. These sedan, it was assumed that the 70 percent aged conversion ef- latter costs were used for the aftertreatment system cost esti - ficiency currently achievable with NSC-based systems would mates in Tables 5.4 and 5.5 because they are considered more be sufficient for emissions compliance through the year representative of the post 2009 period. Obviously, this price 2020. Using the spreadsheet from which the cost estimates situation must be monitored, since it is unlikely to remain at shown in Table 5.6 were obtained, it was also determined that April 2009 levels until 2020. For the sedan with an advanced- for a 2.0-L CI engine for a midsize sedan, the NSC system is level downsized 1.6-L engine, emissions system cost be- a lower cost approach ($688) than is a urea-SCR-based sys- tween November 2007 and April 2009 dropped 30 percent. tem ($837). As a result, Table 5.6 contains no cost estimates Note that the catalyst volumes for the cost computation for for the SCR-urea system for the midsize sedan. This choice the downsized 1.6-L engine were not reduced from the 2.0-L could be changed depending on success in meeting LEV III sizes since the 1.6-L engine must produce the same power TABLE 5.4 Committee’s Estimates of Incremental Cost of CI Diesel Engine over a Baseline SI Gasoline Engine for Replac - ing SI 2.4-L MPFI DOHC Four-Valve Engines in Midsize Sedans (e.g., Malibu, Accord) with Base-Level 2.0-L I4 CI Engines Estimated Cost vs. Baseline 50-State-Saleable ULEV II 2.0-L DOHC CI Diesel Engine Baseline: SI Gasoline 2.4- L MPFI DOHC 4V I4 ($) Common rail 1,800 bar piezo-actuated fuel system with four injectors (@$75), high-pressure pump ($250), fuel rail, 675 regulator, and fuel storage upgrades plus high-energy driver upgrades to the engine control module. Credit for SI content deleted ($32) Variable-geometry turbocharger (VGT) ($250) with electronic controls, aluminum air-air charge air cooler, and plumbing 375 ($125) Upgrades to electrical system: starter motor, alternator, battery, and the 1-kW supplemental electrical cabin heater standard 125 in Europe ($59) Cam, crank, connecting rod, bearing, and piston upgrades, oil lines ($50) plus NVH countermeasures to engine ($40) and 161 vehicle ($71) 215 HP/LP EGR system to suppress NOx at light and heavy loads; includes hot side and cold side electronic rotary diesel EGR valves plus EGR cooler and all plumbing Emissions control system including the following functionality: diesel oxidation catalyst (DOC), catalyzed diesel 688 particulate filter (CDPF), NOx storage catalyst (NSC), EGR catalyst, passive SCR. Stoichiometric MPFI emissions and evaporative systems credit ($245). See Table 5.6 for a detailed breakdown of the emissions control system components leading to the total shown here. On-board diagnostics (OBD) and sensing including an electronic throttle control ($25), four temperature sensors (@$13), 154 wide-range air-fuel ratio sensor ($30), two pressure-sensing glow plugs (@17), two conventional glow plugs (@$3), and Delta-P sensor for DPF ($25). Credit for two switching O2 sensors (@$9). Total variable cost with credits for SI parts removed. Excludes any necessary transmission, chassis, or driveline upgrades. 2,393 NOTE: The credit for downsizing from V6 to I4 included in the Martec Group, Inc. (2008) study was not used in the committee’s estimates since baseline 2007 midsize sedan SI gasoline engines were not V6 but 2.4-L I4 engines. Cost estimates for aftertreatment systems reflect April 2009 prices for platinum group metals.

OCR for page 61
76 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES TABLE 5.5 Committee’s Estimates of Incremental Cost of CI Diesel Engine over a Baseline SI Gasoline Engine for Cost Estimations to Replace SI MPFI DOHC Four-Valve 4.0- to 4.2-L Six-Cylinder Engine in a Midsize Body-on-Frame SUV (e.g., Explorer, Durango) with a 3.5-L V6 DOHC CI Engine Estimated Cost vs. Baseline 50-State-Saleable ULEV II 3.5-L V6 DOHC CI Diesel Engine Baseline: SI Gasoline DOHC 4V 4.0-4.2-L Six Cylinder ($) Common rail 1,800 bar piezo-actuated fuel system with six injectors (@$75), high-pressure pump ($270), fuel rail, 911 regulator and fuel storage upgrades plus high-energy driver upgrades to the engine control module. Credit for MPFI content deleted ($48). Variable-geometry turbocharger (VGT) ($350) with electronic controls, water-air charge air cooler, circulation pump, 485 thermostat/valve and plumbing ($135) Upgrades to electrical system: starter motor, alternator, battery, and the 1.5-kW supplemental electrical cabin heater 167 standard in Europe ($99) Cam, crank, connecting rod, bearing, and piston upgrades, oil lines ($62) plus NVH countermeasures to engine ($47) and 194 vehicle ($85) 226 HP/LP EGR system to suppress NOx at light and heavy loads; includes hot side and cold side electronic rotary diesel EGR valves plus EGR cooler and all plumbing Emissions control system including the following functionality: DOC, CDPF, selective catalytic reduction (SCR), urea 964 dosing system ($363). Stoichiometric MPFI emissions and evaporative systems credit ( $343). See Table 5.6 for a detailed breakdown of the emissions control system components leading to the total shown here. On-board diagnostics (OBD) and sensing including four temperature sensors (@$13), wide-range air-fuel ratio sensor 227 ($30), NOx sensor ($85), two pressure-sensing glow plugs (@17), four glow plugs (@$3), and Delta-P sensor for DPF ($25). Credit for four switching O2 sensors (@$9) Total variable cost with credits for SI parts removed. Excludes any necessary transmission, chassis, or driveline upgrades. 3,174 NOTE: The credit for downsizing from V8 to V6 included in Martec Group, Inc. (2008) was not used here because the baseline 2007 SI engine was a V6, not the V8 assumed in Martec Group, Inc. (2008). Aftertreatment system cost estimates reflect April 2009 prices for platinum group metals. TABLE 5.6 Cost Estimates for Exhaust Emissions Aftertreatment Technologies Capable of Enabling Tier 2, Bin 5 Compliance Midsize Car Midsize Car Midsize SUV Midsize SUV (e.g., Malibu) (e.g., Malibu) (e.g., Explorer), (e.g., Explorer), Catalytic Device Sizing Catalytic Device Sizing Catalytic Device Sizing Catalytic Device Sizing Based on 2 L Based on 2 L Based on 3.5 L Based on 3.5 L Item (Nov. 2007 PGM prices) (Apr. 2009 PGM prices) (Nov. 2007 PGM prices) (Apr. 2009 PGM prices) DOC 1 Monolith and can $52 $52 $52 $52 PGM loading $174 $139 $210 $200 DOC 2 Monolith and can Not used $0 $52 $52 PGM loading Not used $0 $73 $70 EGR catalyst Monolith and can $7 $7 Not used Not used PGM loading $22 $13 Not used Not used Coated DPF Advanced cordierite brick and can $124 $124 $270 $270 PGM loading $160 $131 $29 $26 NSC system Catalyst brick and can $114 $114 Not used Not used PGM loading $533 $314 Not used Not used SCR-urea system SCR brick and can $39 $39 $274 $274 Urea dosing system Passive SCR Passive SCR $363 $363 Stoichiometric gasoline emissions −$245 −$245 −$343 −$343 and evaporative system credit Emissions System Total $980 $688 $980 $964 NOTE: The significant impact of platinum group metals (PGM) commodity prices is illustrated by the difference between the costs in columns 2 and 4 (based on November 2007 prices) and the costs in columns 3 and 5 (based on April 2009 prices).

OCR for page 61
77 COMPRESSION-IGNITION DIESEL ENGINES output as the 2.0-L engine, requiring that exhaust gas flow way down the cylinder to raise the temperature of the gases rates remain virtually unchanged. For the SUV, a smaller 10 by partial oxidation of this regeneration fuel in the cylinder percent emissions system cost drop was observed due to the and completion of oxidation of that fuel in the oxidation lower PGM usage with SCR-urea aftertreatment for out-of- catalyst, some fuel from the high-pressure spray reaches the engine NOx control for the SUV. With SCR-urea systems, cylinder wall and some of that fuel escapes past the piston only the SCR device contains no PGM. As can be observed rings down into the crankcase, where it dilutes the lubricat- from examination of the entries in Table 5.6, DOC1, DOC2, ing oil with fuel. This dilution requires more frequent oil and the coated DPF (called CDPF) all utilize PGM wash- changes to protect engine durability. Since frequency of oil coats. As noted earlier, the spreadsheet used to generate the changes is a marketing attribute, the choice of substrate has aftertreatment cost estimates shown in Table 5.6 is available multiple implications, namely cost, durability, mass, and for recomputing the aftertreatment system cost estimates oil-change interval. should PGM commodity prices change significantly. Advanced cordierite is emerging as a compromise be- Finally, there is a technology choice involved in DPF tween the properties of Si-C and conventional cordierite systems. The four substrate options currently available for (Tilgner et al., 2008). Therefore, for the purpose of this particulate filters are silicon carbide (Si-C), conventional report, it has been assumed that new DPF applications will cordierite, advanced cordierite, and acicular mullite. Con- utilize advanced cordierite (as was assumed for the estimates ventional cordierite is used for most nonparticulate filter in the Martec [2008] report) and that existing Si-C applica- substrates (e.g., DOC and NSC catalysts), whereas Si-C has tions will be converted to advanced cordierite for the next de- been the predominant choice for light-duty DPF usage in sign and development cycle. Thus the cost estimates shown Europe. Conventional cordierite is less expensive and lower in Table 5.7 are based on the use of advanced cordierite for in mass than Si-C. On the other hand, Si-C has much higher DPF monoliths. thermal conductivity and strength, which are very favorable Finally, acicular mullite has recently been introduced to properties for withstanding regeneration without local hot the market. This new material has a number of properties spots causing thermal stress cracking and ultimate failure of that are potentially advantageous for exhaust filtration. First, the filter. As a result of these property differences, Si-C filters this material appears to have lower pressure drop than the are typically filled (i.e., loaded) with about twice the amount other materials due to higher porosity. According to material of particulate (e.g., 8-9 g/L) during vehicle operation before property specifications (Dow, 2009), this higher porosity and regeneration is carried out, whereas conventional cordierite lower pressure drop remain when catalytic coatings are ap- filters must be regenerated after about half that loading (e.g., plied. As a result, it may be possible to integrate additional 4-5 g/L) of particulate. exhaust aftertreatment system components (e.g., combining There are two results from this difference. First, conven- SCR and DPF units into one component), thus reducing tional cordierite-based filter systems tend to require more system cost, packaging volume, and complexity. The first frequent regenerations with associated FC increases. Second, production application of this material is expected in 2011, since during regeneration fuel is injected into the engine after which its technical potential and cost tradeoff relative cylinder during the expansion stroke with the piston part to other materials will become clearer. TABLE 5.7 Comparison of CI Engine Cost Estimates from Different Sources and the Committee’s Estimates I4 CI V6 CI Engine Sizing Aftertreatment System Configurations PGM Cost Dollar Source Engine ($) Engine Methodology Specified and PGM Loadings Basis Basis Martec Group Inc. (2008) 2,361 3,465 Partially Yes Nov. 2007 2007 EPA (2009) 2,052 2,746 Yes Configuration, yes; sizing-loading, no Not specified 2007a Duleep (2008/2009) 1,975 2,590 No Configuration, yes; sizing-loading, no Not specified 2008 2,667 3,733 Partially Assumed to be based on those of Nov. 2007 2007 DOT/NHTSA (2009)b Martec Group, Inc. (2008) 2,393 3,174 Yes Apr. 2009 2007 NRC (2010)c Yesd aEPA 2009 estimates provided were for dollar-year-basis 2002 for engine and 2006 for aftertreatment. The numbers shown have been corrected by apply - ing the ratios of the yearly producer’s price index (1.0169 for 2002 to 2007 and 1.0084 for 2006 to 2007). However, significant technology development has taken place since 2002, and so it is likely that technology-based component specifications and associated costs have changed. bCosts from Tables IV-21, IV-22, and IV-23 of DOT/NHTSA (2009) were divided by 1.5 to convert from RPE (retail price equivalent) to cost. cNRC (2010) refers to the present report. The CI engine costs are for base-level specifications. Detailed breakdowns of the committee’s cost estimates are given in Tables 5.4 and 5.5. dThe spreadsheet used to compute aftertreatment system costs for the present work utilizes the configuration, sizing, and washcoat loadings included in the December 2008 version of the Martec Group, Inc. (2008) study.

OCR for page 61
78 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES Comparison of Cost Estimates with Those of Other $2,671 to $2,379. Finally, an additional pressure-sensing Sources glow plug was added to provide OBD backup for the single pressure-sensing glow plug assumed in the Martec BOM The cost estimates from Martec Group, Inc. (2008), (replace 1 ceramic glow plug @$3 with pressure-sensing EPA (2009), and Duleep (2008/2009) are summarized in glow plug @$17 for net increase of $14). That brought the Table 5.7. From the left, the columns show: present estimate to the $2,393 shown in Tables 5.4 and 5.7. For the SUV case, the Martec analysis assumed that a • The cost estimate source; 3.0-L V6 CI engine would replace a V8 SI engine. As is dis- • The cost estimates for replacing the baseline I4 SI en- cussed above for the I4 case, for the case of a baseline 2007 gines in 2007 model-year midsize sedans (e.g., Malibu, midsize SUV (e.g., Explorer, Trailblazer), the baseline SI Camry) with CI engines; engine was a 4.0- to 4.2-L six-cylinder engine rather than • The cost estimates for replacing the baseline six- the V8 assumed in the Martec analysis. Therefore, the down- cylinder SI engines in 2007 model-year midsize SUVs sizing credit from V8 to V6 used in the Martec analysis ($270) (e.g., Explorer, Trailblazer) with V6 CI engines; was not included for the present analysis, increasing the cost • W hether the sources include details on how the estimate from $3,465 to $3,735. The Martec analysis assumed displacements for the replacement CI engines were a two-stage turbo system for the 3.0-L V6 engine system. chosen; For the comparisons in Table 5.7, only the 3.5-L base-level • Whether the sources include details on exhaust after- technology engine was included to be compatible with the treatment system configurations, component sizing, packages assumed in EPA (2009) and Duleep (2008/2009). and catalyst washcoat loading; Therefore, the air system cost from the Martec analysis was • What is the timing basis for PGM commodity costs; reduced for the present analysis by replacing the two-stage • What is the dollar basis year. turbo system cost estimate ($1,030) with that for a single- stage system ($485). That reduced the estimate from $3,735 Present Cost Estimates Compared to Martec Estimates to $3,190. Finally, the increase in displacement from the Martec 3.0-L displacement to the 3.5 L of the present analysis Although the cost estimates developed in the present along with the use of the April 2009 PGM prices rather than study were based on the estimates from Martec Group, Inc. the November 2007 PGM prices used by Martec reduced the (2008), a number of revisions were made to the Martec aftertreatment system cost from $980 to $964, which in turn estimates. First, the Martec estimates assumed that the 2-L reduced the total V6 SUV replacement cost from $3,190 to four-cylinder CI engine replaced a V6 SI engine in the mid- the $3,174 shown in Tables 5.5 and 5.7. size sedan vehicle. As a result, Martec included a downsizing credit resulting from the savings from the elimination of two Present Cost Estimates Compared to EPA Estimates cylinders and their associated parts. Whether or not it is ap- propriate to include such a credit depends on what baseline The EPA cost estimate shown in Table 5.7 for the I4 CI vehicle is assumed. Because of the timing of the EISA that replacement for the 2.4-L SI engine is $2,052, which is $341 motivated the present study, the baseline vehicles for the less than the committee’s estimate of $2,393. Using detailed present study are 2007 model-year vehicles. The vehicle breakdowns of the EPA estimates (EPA, 2009), one major class that would utilize the 2.0-L CI engine, namely the 2007 difference is the cost credits used in the EPA breakdown for midsize sedan (e.g., Malibu, Camry), typically used a four- parts removed from the SI engine. The EPA estimate for the cylinder 2.4-L SI engine with 4/5-speed automatic transmis- gasoline fuel system removed was $240 ($165 for injectors sion. Therefore, for the present study, the downsizing credit and rail and $75 for fuel pump and vapor recovery (Evap) for reducing the number of cylinders was excluded from the system, whereas that used for the present work from Martec cost estimate since a four-cylinder CI engine would replace Group, Inc. (2008) was $32 for the injection system and a four-cylinder SI engine. This increased the estimate from $37 for the Evap canister and purge valve (included within the Martec value of $2,361 by $310 to $2,671. Second, the the $245 emissions system credit). The fuel pump for the Martec cost estimates were based on November 2007 com- gasoline system is actually replaced by the low-pressure modity prices for the precious metals used in the exhaust supply pump for the CI fuel system, which is very similar aftertreatment system washcoats. Based on the detailed to the gasoline pump, and so there should be no credit for exhaust aftertreatment system specifications provided in the that item. The injectors and rail are extremely high-volume Martec (2008) report, the committee constructed a spread- commodity items sold by suppliers at close to cost because sheet to compute the exhaust aftertreatment system costs, of the strong global competition for such parts. Therefore, the and April 2009 rather than November 2007 PGM prices were $32 credit used for those items is considered representative. used. This change was made to reflect the significant com- The difference between the EPA estimate and the commit- modity price deflation since November 2007. The difference tee’s estimate for the fuel system and vapor recovery is thus amounted to $292, which lowered the cost estimate from $240 versus $69. The EPA assumed a $75 credit for ignition

OCR for page 61
79 COMPRESSION-IGNITION DIESEL ENGINES system parts removed from the SI engine. The pencil coils market place, although the benefits of this technology will used in 2007 ignition systems are again extremely high- not show up on the EPA certification tests because those are volume commodity items. The ignition control drivers used conducted using standardized certification fuels for which in such systems are up-integrated into the ECM, and so the engines are calibrated during development. there is effectively no savings from their removal. For the CI As shown in Table 5.7 for the V6 midsize SUV case, engine, a glow plug and wire is required for each cylinder, the EPA estimate for replacing the SI engine with a CI en- so the SI to CI ignition cost difference was considered $0. gine was $2,746, which was $694 greater than that for the There were other differences in the individual item estimates I4 CI engine substitution. The corresponding increment as between the EPA estimate and that from the present estimate determined in the present study was $781. The differences as well. The EPA estimate for the turbocharger system was between the detailed items in the two cost estimates remain less than that of the present study ($181 versus $375). The similar to those already discussed for the I4 case, and since EPA estimate for emissions controls appeared to reflect a the total cost differences were similar, the details are not somewhat different approach to emission control, with more discussed here. However, for the V6, both estimates assumed emphasis on aftertreatment and less emphasis on in-cylinder the urea-SCR approach for NOx aftertreatment. combustion-based control of emissions. This approach is il- lustrated by the EPA choice of a urea-SCR strategy for NOx Present Cost Estimates Compared to EEA (Duleep) aftertreatment while that for the present approach was an Estimates NSC-based approach. The present approach also included an HP/LP EGR system, whereas the EPA system did not. The The EEA (Duleep, 2008/2009) variable cost estimate HP/LP EGR system will lower engine-out emissions, where- for replacing the 2.4-L SI engine with a 2.0-L CI engine as the NSC NOx conversion efficiency is lower than that of (Table 5.7) was $1,975. This total consisted of $1,145 for the urea-SCR approach, as noted earlier in the discussion of the engine and $830 for emissions control. The present NOx aftertreatment system technologies. As a result, the EPA study’s engine cost estimate was $1,336. One of the larger emissions system cost estimate was significantly higher than differences between these two estimates was for the turbo that from the present work ($1,220 versus $903 ($688 for system—EEA estimated a total of $280 and the Martec- aftertreatment plus $215 for HP/LP EGR)). The urea-SCR based present study’s estimate was $250 for the VGT turbo subsystem cost in the EPA estimate versus that for the NSC with electronic controls and $125 for the intercooler and in the present study was $670 versus $428, and the EPA plumbing, for a total turbo system cost of $375, or $95 above CDPF cost was estimated at $480 versus $255 for the pres- the EEA estimate. Also, the EEA estimate did not include ent study. No information was available concerning CDPF a cabin heater, which is standard with CI diesel vehicles substrate volume or PGM loading to understand the source of and which Martec estimated at $59. For exhaust emissions these differences in more detail. The present study assumed control, the differences between the EEA estimates and the that the aftertreatment system would also require an EGR Martec-based estimates used in the present study were also catalyst ($20) to control EGR cooler fouling, and a passive significant. EEA assumed an integrated DPF and NSC unit SCR catalyst ($39), which would provide a small amount (called DPNR), which is proprietary to Toyota. All other of NOx reduction on the US06 test using the small amount of OEMs are using separate DPF and NSC units. The EEA esti- ammonia produced by the NSC at the higher load conditions mate assumed $730 for the DPNR unit, but no cost basis was of the US06 test rather than urea from a separate system like specified for the PGM prices or loadings. The present study that in the urea-SCR system. OEMs will make the choice of assumed $688 (see Table 5.6) based on April 2009 PGM emissions control strategy based on many factors, including prices for the separate DPF and NSC units. EEA assumed cost, durability, customer convenience, and packaging. In $60 for the EGR system and cooler, whereas the present addition to cost differences, the urea-SCR approach requires study estimated $215 for an HP/LP EGR system (for details finding space to package a urea supply tank, which is more see Table 5.4). As noted in earlier discussion of emissions problematic in a smaller vehicle like the midsize sedan than control technology, a combined HP/LP EGR system has for a larger vehicle like an SUV. As noted earlier, the 2009 many advantages for reducing engine-out NOx, thus reduc- VW Jetta utilizes a system very much like the system as- ing the NOx conversion requirements for the aftertreatment sumed in the present study. The other area in which different system. The LP EGR system requires several control valves components were assumed by the EPA was for OBD and and cooler in addition to those for the HP EGR system. The sensing. The present study assumed four temperature sen- 2009 VW Jetta has such an HP/LP EGR system. For oxida- sors ($52) and two pressure-sensing glow plugs ($34), which tive cleanup of the exhaust (e.g., unburned HC, CO, and were not included in the EPA system. As noted earlier in soluble particulates), a DOC (diesel oxidation catalyst) is discussions about combustion technologies, the closed-loop used. EEA assumed $50 for the DOC. Again, no informa- cylinder-pressure sensing system is beneficial for minimiz- tion was provided about volume, PGM loading, or PGM cost ing engine fuel consumption and emissions when different basis for the EEA estimate. The present study assumed $52 fuels of widely different cetane ratings are encountered in the for the monolith and housing and $139 for the PGM wash-

OCR for page 61
80 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES coat cost based on April 2009 PGM prices. The emissions estimate from $3,174 to $3,719, leaving a difference between control system cost estimate differences then totaled $227. the NHTSA estimate and the present estimate of about $14. For the V6 SUV case, the EEA estimate was $2,590, There are also other differences between the assump- whereas that of the present study was $3,174. The EEA esti- tions made in the present study and those of the Martec mate for the engine was $1,715 versus $1,983 for the present study. For the engine sizing methodology used herein, the study. Of the $268 difference, the majority is explained by baseline six-cylinder engine for the midsize vehicle class of the lack of a cabin heater in the EEA estimate and inclusion about 4.2 L downsized by the assumed 83 percent is 3.5 L, of the cabin heater for the present study at $99 (more costly whereas the Martec study assumes 3.0 L. According to the than that of the midsize sedan I4 vehicle because of the larger costing methodology used in the present study, the increase cabin volume for the midsize SUV with the V6) and the air of displacement from 3.0 L to 3.5 L increases cost (entirely system (turbocharger and intercooler) for which EEA esti- as a result of aftertreatment systems cost) from $921 to $964. mated $365 versus $485 for the present study. The remainder Subtracting this difference from the engine cost estimate of of the difference was due to emissions control. Again, one of $3,174 increases the cost differential between the NHTSA the main differences was the use of an HP/LP EGR system estimate and the present study from $14 to $57. As for the for the present study as included in the Martec BOM but not remaining difference, there is insufficient information in in the EEA estimate ($86 difference). In addition, the present the NHTSA report to understand the sources of this differ- study included the use of a second DOC ($122) included in ence, although it is less than 10 percent, which is well within the Martec BOM that was worked out in collaboration with the uncertainty of these cost estimates in general. OEMs and suppliers. Advanced-Level CI Engine Cost Estimates Present Cost Estimates Compared to NHTSA Estimates Cost estimates for the technologies necessary to raise According to the NHTSA final ruling for 2011 (DOT/ base-level CI engines to advanced-level engines inherent in NHTSA, 2009), costs for CI engines and DCT6 transmis- the gains described in Table 5.2 are listed in Table 5.8. sions were also derived from the Martec estimates. For the 2.0-L I4, the NHTSA number from Table 5.7 is $2,667, Advanced-Level Transmission Cost Estimates whereas the corresponding number from the present study is $2,393. Most of the difference between these estimates There seems to be an emerging consensus that dual-clutch is due to the $292 reduction in aftertreatment system costs automatically shifted manual transmissions (DCTs) offer a used in the present study and derived from using April 2009 very attractive combination of efficiency and driver satisfac- PGM prices rather than the November 2007 prices reflected tion with acceptable cost. In the Ricardo, Inc., FSS studies in the Martec numbers presumably used by the NHTSA. for the EPA (EPA, 2008), CI engines were combined with It is not known whether the NHTSA estimate includes the DCT6 units for the simulations, as noted in earlier discus- downsizing credit or not. sions of Table 5.1. For that reason, it was assumed for the The NHTSA cost estimate of $5,600 retail price equiva- present analysis that the CI replacements for SI engines lent ($3,733 cost) from Tables IV-21, IV-22, and IV-23 would use DCTs. Transmission technologies are discussed (DOT/NHTSA, 2009) for the larger vehicle classes (e.g., in Chapter 7, which considers non-engine vehicle technolo- large car versus subcompact, compact, and midsize car) is gies. Cost estimates for advanced transmissions used for this assumed to derive from the Martec cost estimate of $3,465 committee’s work are also shown there and are summarized for V6 diesel (Martec Group, Inc., 2008, p. 37). The corre- in Table 7.10. sponding value for the V6 CI engine from the present study was $3,174. A significant portion of the $559 difference Summary of Total SI to CI Power Train Replacement Cost between the NHTSA estimates and those of the present work Estimates is due to the inclusion in the Martec, and presumably also in the NHTSA, estimates of two-stage turbocharger systems The total estimated costs to replace 2007 model-year SI that for the present study correspond to advanced-level en- power trains with base-level and advanced-level CI power gine technology, as described in the section “Engine Sizing trains for the example midsize sedan and midsize SUV Methodology.” As noted above, the costs from the present vehicles indicated in Tables 5.4 and 5.5 are summarized in work that were used in Table 5.7 were those for the base- Table 5.9. level technology configuration. The base level was assumed to use single-stage VGT turbo systems and the advanced FINDINGS level to use two-stage turbo systems. The cost estimate from the present work, which is included in Table 5.7, is for the Based on a combination of analysis and engineering judg- base-level CI engine. Including the two-stage turbo system in ment applied to information collected from many sources, the cost estimate from the present study would increase the the committee’s key findings are as follows regarding tech-

OCR for page 61
81 COMPRESSION-IGNITION DIESEL ENGINES TABLE 5.8 Committee’s Estimates of Incremental Costs to Implement Advanced-Level Diesel Developments (downsizing, thermodynamic improvements, friction reduction, and engine accessory improvements) Whose Estimated Potential for Reducing Fuel Consumption Is Summarized in Table 5.2 Midsize Car Midsize SUV (e.g., Malibu) (e.g., Explorer) Item 1.6-Liter I4 2.8-Liter V6 Comment Downsize engines from 2-L I4 to 1.6-L $50 $75 Higher load capacity rod bearings and head gasket for higher I4 and from 3.5-L V6 to 2.8-L V6 cylinder pressures (~$12.50/cylinder) Two-stage turbocharger system $375 $545 Additional air flow control valves, piping, cost of additional turbo, water-to-air intercooler with separate pump, control valve Dual-pressure oil pump $5 $6 Switchable pressure relief valve for high or low oil pressure Non-recirculating low-pressure (LP) $10 $12 Variable output LP pump controlled by high-pressure (HP) pump fuel pump output Cylinder pressure sensors — — Two pressure-sensing glow plugs, one to sense fuel property differences, second to provide on-board diagnostics durability backup for first, already included for both I4 and V6 in Tables 5.3 and 5.4 Low-pressure exhaust gas recirculation — $95 Additional piping (~$20) and valves (e.g., integrated back pressure (EGR) and LP EGR rate ~$75), much more difficult to package for V6 engine with underfloor diesel particulate filter, cost for I4 already included in Table 5.4 Direct-acting HP (maximum injection $80 $120 $20/injector, benefits derived from combination of higher rail pressures >2,000 bar) piezo injectors pressure and more injector controllability Total $520 $853 TABLE 5.9 Estimated Total Costs to Replace 2007 Model-Year SI Power Trains with Base- and Advanced-Level CI Power Trains for Example Midsize Sedan and Midsize SUV-Type Vehicles Base-Level CI Engine Advanced-Level CI Engine Midsize Sedan I4 engine $2,393 (Table 5.4) or $2,913 (Tables 5.4 and 5.8) or $2,400 (when rounded to nearest $50) $2,900 (when rounded to nearest $50) $140-$400 (Table 7.10) $140-$400 (Table 7.10) DCT6/7a transmission Total $2,550-$2,800 (when rounded to nearest $50) $3,050-$3,300 (when rounded to nearest $50) Midsize SUV V6 engine $3,174 (Table 5.5) or $4,027 (Tables 5.5 and 5.8) or $3,150 (when rounded to nearest $50) $4,050 (when rounded to nearest $50) DCT6/7 transmission $140-$400 (Table 7.10) $140-$400 (Table 7.10) Total $3,300-$3,550 (when rounded to nearest $50) $4,150-$4,450 (when rounded to nearest $50) aNote that the higher of the two estimates shown in Table 7.10 is for a 6/7-speed dual-clutch transmission (DCT). In accordance with the potential fuel consumption reduction gains discussed in Table 5.2 due to transmission improvements, it was assumed that 7-speed versions would be used. Due to the wide range of cost estimates for DCTs as discussed in Chapter 7, no adjustment was made for the higher torque requirements of the V6 CI. nology combinations for reducing the fuel consumption of (DCTs) (6-speed) and more efficient accessories packages 2007 model-year SI gasoline engine vehicles by equipping can reduce fuel consumption by an average of about 33 them with advanced CI diesel power trains. percent (or reduce CO2 emissions by about 23 percent) on an equivalent vehicle performance basis. Advanced-level Finding 5.1: By a joint effort between OEMs and suppliers, CI diesel engines with advanced DCTs could reduce fuel new emissions control technology has been developed to consumption by about an additional 13 percent for larger ve- enable a wide range of light-duty CI engine vehicles to meet hicles and by about 7 percent for small vehicles with engine the 2010 Tier 2, Bin 5, LEV II emissions standards. displacements less than 1.5 L. Finding 5.2: Replacing 2007 model year MPFI SI gaso- Finding 5.3: The characteristics of CI diesel engines that en- line power trains with base-level CI diesel engines with able their low fuel consumption apply over the entire vehicle advanced dual-clutch (automated manual) transmissions operating range from city driving to highway driving, hill

OCR for page 61
82 ASSESSMENT OF FUEL ECONOMY TECHNOLOGIES FOR LIGHT-DUTY VEHICLES climbing, and towing. This attribute of CI diesel engines is Dow. 2009. Available at http://www.dow.com/PublishedLiterature/ dh_02df/0901b803802df0d2.pdf? filepath=automotive/pdfs/ an advantage when compared with other technology options noreg/299-51508.pdf&fromPage=GetDoc. that are advantageous for only part of the vehicle operating Duleep, K.G. 2008/2009. Diesel and hybrid cost analysis: EEA versus range (e.g., hybrid power trains reduce fuel consumption Martec, presentation made to NRC Committee, February 25, 2008, primarily in city cycle/city driving). updated June 3, 2009. EIA (Energy Information Administration). 2009a. Light-Duty Diesel Vehicles: Efficiency and Emissions Attributes and Market Issues. Feb- Finding 5.4: The identified advanced-level technology ruary. Available at http://www.eia.doe.gov/oiaf/servicerpt/lightduty/ improvements to CI diesel engines are expected to reach execsummary.html. market in the 2011-2014 time frame, when advanced tech- EIA. 2009b. Diesel Fuel Prices. Available at http://tonto.eia.doe.gov/oog/ nology additions to SI gasoline engines will also enter the info/gdu/gasdiesel.asp. Accessed May 9, 2009, and June 5, 2009. market. Thus, there will continue to be a fuel consumption EPA (U.S. Environmental Protection Agency). 2005. Document 420-F- 05-001. Available at http://www.epa.gov/otaq/climate/420f05001.htm. and cost competition between these two power train systems. EPA. 2008. A Study of Potential Effectiveness of Carbon Dioxide Reducing For the period 2014-2020, further potential fuel consump- Vehicle Technologies. Report 420r80040a. Revised June. tion reductions for CI diesel engines may be offset by fuel EPA. 2009. Updated cost estimates from those in U.S. EPA, 2008. Commit - consumption increases due to engine and emissions system tee e-mail communications with EPA, May 27 and May 28. changes required to meet stricter emissions standards (e.g., Hadler, J., F. Rudolph, R. Dorenkamp, H. Stehr, T. Düsterdiek, J. Hilzendeger, D. Mannigel, S. Kranzusch, B. Veldten, M. Kösters, and A. Specht. 2008. LEV III). Volkswagen’s new 2.0 L TDI engine fulfills the most stringent emission standards, 29th Vienna Motor Symposium. Finding 5.5: CI diesel engine market penetration will be Iwabuchi, Y., K. Kawai, T. Shoji, and Y. Takeda. 1999. Trial of new concept strongly influenced both by the incremental cost of CI diesel diesel combustion system—Premixed compression-ignition combus- power trains above the cost of SI gasoline power trains and tion. SAE Paper 1999-01-0185. SAE International, Warrendale, Pa. Joergl, Volker, P. Keller, O. Weber, K. Mueller-Haas, and R. Konieczny. by the price differential of diesel fuel relative to gasoline. 2008. Influence of pre-turbo catalyst design on diesel engine perfor- The estimated incremental cost differential for base-level and mance, emissions, and fuel economy. SAE Paper 2008-01-0071. SAE advanced-level I4 CI diesel engines to replace 2007 model- International, Warrendale, Pa. year midsize sedan SI gasoline engines ranges from $2,400 Kanda, T., T. Hakozaki, T. Uchimoto, J. Hatano, N. Kitayama, and H. Sono. (base level) to $2,900 (advanced level). For base-level I4 2005. PCCI operation with early injection of conventional diesel fuel. SAE Paper 2005-01-0378. SAE International, Warrendale, Pa. engines combined with DCTs, power train replacement cost Keller, P.S., V. Joergl, O. Weber, and R. Czarnowski. 2008. Enabling com- is estimated at $2,550 to $2,800 and for advanced-level I4 ponents for future clean diesel engines. SAE Paper 2008-01-1530. SAE power trains is estimated at $3,050 to $3,300 (both rounded International, Warrendale, Pa. to the nearest $50). For midsize 2007 model-year SUVs, the Martec Group, Inc. 2008. Variable Costs of Fuel Economy Technologies. estimated cost for replacement of SI gasoline engines with Prepared for Alliance of Automobile Manufacturers, June 1; amended September 26 and December 10. base-level and advanced-level V6 CI diesel engines ranges Mattes, Wolfgang, Peter Raschl, and Nikolai Schubert. 2008. Tailored from $3,150 (base level) to $4,050 (advanced level) (both DeNOx concepts for high-performance diesel engines. Second MinNOx rounded to the nearest $50). For V6 CI engines combined Conference, June 19-20, Berlin. with DCTs, the estimated V6 CI power train replacement Müller, W., et al. 2003. Selective catalytic reduction—Europe’s NOx reduc- cost increment over 2007 model-year SI power trains is tion technology. SAE 2003-01-2304. SAE International, Warrendale, Pa. Myoshi, N., et al. 1995. Development of new concept three-way catalyst $3,300 to $3,550 (base level), and the advanced-level power for automotive lean-burn engines. SAE Paper 95809. SAE International, train incremental cost is $4,200 to $4,500 (both rounded Warrendale, Pa. to nearest $50). These costs do not include the retail price NRC (National Research Council). 2002. Effectiveness and Impact of Cor- equivalent factor. porate Average Fuel Economy (CAFE) Standards. National Academy Press, Washington, D.C. Peckham, John. 2003. How JD Power/LMC calculate 16% light-duty diesel REFERENCES sales share for North America. Diesel Fuel News, October 13. Pickett, L.M., and D.L. Siebers. 2004. Non-sooting, low flame temperature Bression, G., D. Soleri, S. Savy, S. Dehoux, D. Azoulay, H.B-H. Hamouda, mixing-controlled DI diesel combustion. SAE Paper 2004-01-1399. L. Doradoux, N. Guerrassi, and N. Lawrence. 2008. A study of methods SAE International, Warrendale, Pa. to lower HC and CO emissions in diesel HCCI. SAE Paper 2008-01- Ryan, T.W., and T.J. Callahan. 1996. Homogeneous charge compres- 0034. SAE International, Warrendale, Pa. sion ignition of diesel fuel. SAE Paper 961160. SAE International, Diesel Forum. 2008. Available at http://www.dieselforum.org/DTF/ Warrendale, Pa. news-center/pdfs/Diesel%20Fuel%20Update%20-%20Oct%202008.pdf. Styles, D., J. Giuliano, J. Hoard, S. Sluder, J. Storey, S. Lewis, and M. DieselNet. 2008. February 22. Available at http://www.dieselnet.com/ Lance. 2008. Identification and control of factors that affect EGR cooler news/2008/02acea.php. fouling. 14th Diesel Engine-Efficiency and Emissions Research Confer- DOT/NHTSA (Department of Transportation/National Highway Traffic ence, Dearborn, Mich. S afety Administration). 2009. Average Fuel Economy Standards Tilgner, Ingo-C., T. Boger, C. Jaskula, Z.G. Pamio, H. Lörch, and S. Gomm. Passenger Cars and Light Trucks—Model Year 2011. Docket No. 2008. A new diesel particulate filter material for passenger cars: Cordi - NHTSA-2009-0062, RIN 2127-AK29, March 23. Washington, D.C. erite diesel particulate filters for the new Audi A4 V6 TDI, 17. Aachener Kolloquium Fahrzeug- und Motorentechnik, p. 325.

OCR for page 61
83 COMPRESSION-IGNITION DIESEL ENGINES ANNEX Table 5.A.1 shows the data used in Figure 5.4 for the percentage reduction of fuel consumption in 2009 European vehicle platforms offered with both SI gasoline engines and CI diesel engines in configurations that provide virtually equal performance (i.e., 0 to 100 km/h acceleration times within 5 percent between SI and CI). TABLE 5.A.1 Data Used in Figure 5.4 Vehicle % FC Reduction Audi A3 30.88 BMW 520 25.00 Dodge Avenger 20.51 Ford Fiesta 26.32 Ford Galaxy 36.73 Honda Civic 21.21 Honda CR-V 18.52 Jaguar XF 29.25 Mercedes E230 31.18 Mercedes S350 17.65 Toyota Yaris 25.00 Toyota RAV4 24.42 VW Jetta 28.38 Peugeot 308 28.79 Renault Laguna 34.62 Audi A8 21.30 Audi �7 18.38 Audi A6 18.18 Mercedes Viano 23.53 AVERAGE 25.25