NOX Emission Controls for Heavy-Duty Vehicles: Toward Meeting a 1986 Standard (1981)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Chapter 3 HEAVY-DUTY GASOLINE ENGINES INTRODUCTION Heavy-duty and light-duty gasoline engines are often similar in design and construction. Some manufacturers, in fact, use the same engines (with modifications) in both heavy- and light-duty vehicles. However, a significant difference between the two applications is the greater fraction of high-power use typical of heavy-duty service. Because light-duty and heavy-duty engines are often similar, basic emis- sion control techniques are common to the two applications. In the past, emission control requirements for light-duty engines have been more stringent than those for heavy-duty engines of the same model year. For this reason, the control technology used on current light-duty engines is a useful source of information on techniques for meeting future, more stringent regulations on heavy-duty engine emissions. Control Techniques Emission controls for gasoline engines are based on engine modifications and/or the addition of exhaust-treatment devices such as catalytic converters. Engine modifications include changes in the air/fuel ratio and spark timing; redesign of combustion chambers, intake manifolds, and fuel metering systems; changes in valve timing; and the addition of exhaust gas recir- culation (EGR). These modifications have been used by themselves to meet moderately stringent emissions standards. However, as standards become more stringent this approach typically results in deteriorating fuel economy and engine performance. As a consequence, catalytic controls are now used in light-duty vehicles to meet the more stringent emission limitations while at the same time allowing the engines themselves to be optimized for fuel economy and performance. Noncatalytic after-treatment of emissions, by addition of a pump to inject air into the hot exhaust manifold, has been used in the past to 47

48 reduce the hydrocarbon and carbon monoxide emissions of light-duty vehicles and is currently used on heavy-duty gasoline engines. This type of emission control can be enhanced by redesigning the exhaust manifold to provide longer residence times at high temperature. Some noncatalytic oxidation of hydrocarbons and carbon monoxide may also be used in combination with catalytic control systems. The use of catalytic converters for emission controls on gasoline-powered light-duty vehicles is virtually universal in the United States. Some vehicles use oxidizing catalysts, which control only hydrocarbon and carbon monoxide emissions, and control NOX emissions by engine modificaticras. Most catalytic systems on l98l light-duty vehicles, however, use three-way catalysts (TWC), which simultaneously control hydrocarbons, carbon monoxide, and NOX. The operation of these three-way systems requires close control of air/fuel ratios. This is achieved by the use of exhaust gas oxygen sensors and electronic feed- back-controlled fuel metering systems. Both carburetors and fuel injection can be used for feedback-controlled fuel metering. The electronic control systems, which are based on microcomputer technology, are capable of monitoring engine variables and ambient conditions such as barometric pressure. The results of these measurements are combined and used to set air/fuel ratio, spark timing, and emissions control system parameters to control emissions while maintaining the best possible performance and fuel economy. EFFECTS OF ENGINE MODIFICATIONS ON EMISSIONS AND FUEL ECONOMY The effects of various engine design features on gasoline engine emissions and performance have been studied for several years. General descriptions of these effects may be found in several references (Patterson and Henein, l972; Obert, l968; National Research Council, l974). A brief summary is given under the following heads. Air/Fuel Ratio Figure 7 shows the effects of variations in air/fuel ratio on emissions, power, exhaust temperature, and fuel economy. Engines without emission controls (especially heavy-duty engines) are typically calibrated to run rich, to ensure good driveability and high power output. Operation at lean air/fuel ratios, which could provide better fuel economy, has often been avoided to minimize performance and driveability problems. Spark Timing Spark timing can be retarded (set to occur closer to the end of the compression stroke) to reduce emissions of hydrocarbons and NOX. However, this lowers fuel economy and performance. Exhaust Gas Recirculation Introducing exhaust gases into the intake charge lowers peak combustion

49 Stoichiometric Air/Fuel Ratio Exhaust „ Temperature Rich Lean AIR/FUEL RATIO Figure 7 Schematic illustration of the dependence of emissions and other engine parameters on air/fuel ratio. Vertical scale is linear; the curves show the relative variation in the indicated parameters for a typical engine.

50 temperatures, which in turn yields lower NOX emissions. Effects on other pollutant species and fuel economy depend on the amount of EGR used and the design of the engine. At low EGR rates, there is little or no effect on these variables. At the high EGR rates required for stringent NOX control without catalysts, fuel economy and performance can deteriorate significantly. The use of EGR at wide-open throttle settings reduces maximum engine power. Valve Timing Modified valve timing, particularly valve overlap (the time interval in which both the intake and the exhaust valve are open simultaneously) provides some internal EGR for NOX control. Combustion Chamber Redesign Reduction of crevices and changes in the surface-to-volume ratio are used to reduce emissions. More sophisticated changes, aimed at better tolerance of EGR, are the subject of current research. Compression Ratio Changes Compression ratios are determined generally by the octane rating of the gasoline available. Decreased compression ratios may provide lower emissions of hydrocarbons and NOX. However, these benefits are usually accompanied by a fuel economy penalty. Fuel System Modifications In addition to the overall air/fuel ratio, the cylinder-to-cylinder variance in air/fuel ratio, as influenced by induction system design, is an important consideration. Although the overall air/fuel ratio of an engine may be pre- cisely controlled, the way fuel and air are mixed in the intake manifold can impart to each cylinder a different air/fuel ratio. In engines without emission controls, the overall ratio is usually set rich enough to ensure that the leanest cylinder will have a ratio rich enough to provide adequate perform- ance. Modifications to the intake system that improve the cylinder-to-cylinder uniformity of the air/fuel ratio allow the overall ratio to be set leaner with- out impairing other engine performance parameters. Such modifications are used to some extent on current heavy-duty engines and to a much greater extent on light-duty engines. PERFORMANCE OF CURRENT HEAVY-DUTY GASOLINE ENGINES Correlation of Emission Data from the Steady-State and Transient Test Cycles Emissions of current production engines are determined by testing on the nine-mode steady-state test cycle. Some data on the emissions performance of

5l l979 model year gasoline engines, from both the new transient cycle and the steady-state cycle, are available. These data were obtained by Southwest Research Institute under contract to EPA (California Air Resources Board, l98l). All data were obtained from engines with noncatalytic control systems. Comparisons of emissions for the two test cycles are shown in Figures 8-l0. The lines shown on each graph are least-squares linear fits of the data. The corresponding equations and correlation coefficients for the l2 engines tested are given below, for hydrocarbons (HC), carbon monoxide (CO), and NOX: Transient HC - 2.26 + 0.553 x (Steady-State HC) (correlation coefficient, 0.460) Transient CO - 68.9 + 0.3l x (Steady-State CO) (correlation coefficient, 0.l2) Transient NOX = l.44 + 0.74 x (Steady-State NOX) (correlation coefficient, 0.73). The poor correlation of the two test cycles for hydrocarbons and carbon monoxide is obvious. The correlation coefficients for these species are not significantly different from zero by a one-tailed t-test at the 0.05 level. The correlation for NOX is somewhat better, with a correlation coefficient significantly different from zero at the 0.005 level. Nevertheless, data on some engines are far from the correlation line. In addition, none of the data used in obtaining this correlation were at low emissions levels; the lowest NOx levels were 3.95 g/bhp-h (steady-state) and 4.29 g/bhp-h (transient). These two lowest values were not obtained with the same engine. Also, these data represent relatively new engines that had been run on a dynamometer for l25 hours as certification engines. It is not clear how the correlation between the two test cycles would change if it were run on engines at the ends of their useful lives, at which point the NOX standard will apply. Although the above correlation equation for NOX is an estimate at best, it does give a rough indication of how an engine with a given NOX emission level on the steady-state cycle will perform on the transient cycle. Examples of Current-Production Low-NOx Engines Table l7 shows data for Chrysler heavy-duty gasoline engines for the l98l model year. These engines are designed to be used in the lowest range of heavy-duty applications (gross vehicle weight ratings less than ll,000 Ib). They are of particular interest because they use oxidation catalysts (with air pumps) to control emissions of hydrocarbons and carbon monoxide. Spark retard and EGR are used for NOX control. The oxidation catalyst allows the

52 -C a £ V) O O m ec < o O oc o X LU O I- z UJ (A • 2 5* »3 • 9 •6 • 4 • 11 • 12 0.5 1.0 1.5 2.0 2.5 3.0 STEADY-STATE CYCLE HYDROCARBON EMISSIONS (g/bhp-h) Figure 8 Comparison of hydrocarbon emission data taken with the transient and steady-state test cycles. The numbers by the data points are arbitrary engine identifiers, used also in Figures 9 and l0 (California Air Resources Board, l98l).

53 x o 125 100 '•6 • 2 o GO oc < o 75 12* 50 25 I I I J 0 10 20 30 40 50 60 STEADY-STATE CYCLE CARBON MONOXIDE EMISSIONS (g/bhp-h) Figure 9 Comparison of carbon monoxide emission data taken with the transient and steady-state test cycles. Numbers by the data points identify the same engines as in Figures 8 and l0 (California Air Resources Board, l98l).

54 10.- Q. I V) V) V) UI X O z o u 55 2 oc 11 •3 I I I I "024 6 8 10 STEADY-STATE CYCLE N0X EMISSIONS (g/bhp-h) Figure l0 Comparison of NOX emission data taken with the transient and steady-state test cycles. Numbers by the data points identify the same engines as in Figures 8 and 9 (California Air Resources Board, l98l).

55 Table l7 Current Emission Levels of Chrysler Heavy-Duty Engines, as Measured by Steady-State Test Procedure Emissions (g/bhp-h)a Catalyst Engine Designation Fuel Hydro— Carbon Consumption carbons Monoxide NOX g/bhp-h Maximum Power hp No 360-l 0.40 l7.06 5.22 3l8 l90 No 36 0-3 b 0.48 8.ll 5.l3 324 l30 Yes 360-l 0.ll l0.60 l.78 406 l90 Yes 360-l 0.09 3.90 4.04 343 l90 Yes 360-3 0.l8 l4.82 l.87 378 l80 Dual 360-3 0.05 ll.90 2.32 374 200 aEmissions values include deterioration factors measured for l,500-hour dyn amorae t e r run. bThe 360-3 engine is designed for more severe use than the 360-l engine. SOURCE: Informal communication, Chrysler Corporation, April l, l98l.

56 engine to be calibrated rich for better driveability, without sacrificing control of carbon monoxide and hydrocarbons emissions. With catalysts, these engines have very low emissions—significantly lower than the California heavy-duty standards of l, 25, and 6 g/bhp-h for hydrocarbons, carbon monoxide, and the sum of hydrocarbons and NOX, respectively. These very low emissions do carry a significant fuel consumption penalty, however. The catalyst-equipped engines consume 8-l7 percent more fuel than their counterparts without catalysts, mainly as a result of the richer air/fuel ratio. Note the accompanying reductions in NOX emissions of the catalyst- equipped engines. The data in Table l7 are from the steady-state test cycle. The correla- tion equation of the previous section, though not validated at these low levels of NOX emissions, can be used to estimate the NOX levels these engines would attain on the transient cycle. The lowest steady-state NOX number(s) in Table l7 then correspond to a value of about 2.8 g/bhp-h on the transient procedure. Chrysler representatives contacted by the committee were not aware of any durability problems with the oxidation catalysts on their engines. They reported no problems with operation of the catalyst at 90 percent of full load, a required test point on the steady-state cycle. Another low-emission engine in current production is the International Harvester 345-cubic-inch-displacement engine. This engine had emissions of 0.3, l8, and 2.6 g/bhp-h respectively for hydrocarbons, carbon monoxide, and the sum of hydrocarbons and NOX, with NOX emissions of 2.3 g/bhp-h on the steady-state cycle (corresponding to an estimated 3.l g/bhp-h on the transient cycle, according to the correlation equation). The engine is rated at l62 hp at 3,600 rpm. The International Harvester engine uses a rich mixture, rather than EGR, to lower NOX emissions. Two air pumps, one for each exhaust bank, are used to control the carbon monoxide emissions that result from rich operation. International Harvester representatives estimated the highway fuel economy loss in going from the l979 model year engine to this current version at about 7 percent. The l979 engine was certified at 5.6 g/bhp-h NOX emissions on an earlier version of the steady-state cycle. Details of this earlier cycle were modified for the l980 year model. If the l979 engine were tested on the latest version of the steady-state cycle, NOX emissions would probably be higher. Thus, it is not possible to make a direct comparison of the fuel economy changes with specific emissions levels. The data on the Chrysler and International Harvester engines reported here are not typical of the current mix of production engines. They do, however, indicate the emissions levels that can be achieved with current heavy-duty control technology. Unfortunately, no transient cycle data were available for these engines. Also, it is not known how the engines would perform at the ends of their useful lives, when the proposed new regulations would apply.

57 EMISSIONS DATA FOR PROTOTYPE l986 CONTROL SYSTEMS Emissions test data relevant to the proposed l986 standards are extremely scarce at present. First of all, the recent implementation of a new heavy-duty transient test procedure involves significant installation expense and effort. In addition, the proposed NOX standards have only recently been announced. Limited test data are available for experimental emissions control systems on essentially new engines. However, at present there appears to be no information on the durability of such systems in the laboratory or in use. Since the heavy-duty gasoline vehicle represents a new and potentially severe application of catalytic emission control technology, the durability question is vital to evaluating the feasibility of meeting proposed standards. Noncatalytic NOX Controls Figure ll represents schematically the systems used for noncatalytic NOX emissions control. As explained earlier, these can include recalibra- tion of air/fuel ratios, rescheduling of ignition timing, and introduction of EGR. Since the severity of the proposed l986 NOX standard virtually dictates the use of catalytic control techology, little information has been developed on the performance of noncatalytic systems at these low NOX levels. Such data, if available, would be helpful in determining minimum feasible emissions levels for noncatalytic systems and the magnitudes of the fuel economy and performance penalties involved. While test data are limited, estimates of the potential effectiveness of EGR coupled with changes in engine calibration, extracted from material provided by the General Motors Corporation (l98l) are presented in Table l8. These data are based on laboratory engine tests and therefore don't address possible impacts on performance or driveability. Transient test cycle fuel economy values are included, however. The data in Table l8 were derived from tests run on a prototype 350-cubic- inch-displacement V-8 engine equipped with dual oxidizing converters to meet l984 hydrocarbon and carbon monoxide standards. Hydrocarbon and carbon monoxide measurements taken with fresh catalyst are well within the standards. With application of EGR, an NOX emission reduction approaching 50 percent of base engine emissions was realized. The lowest recorded NOX emission rate was about 3 g/bhp-h. At this emission level, fuel consumption appeared to increase by about 4 percent compared with base engine levels. However, the statistical significance of this increase is not known. A second estimate of the potential effectiveness of EGR and its impact on fuel consumption can be obtained from data published recently by EPA (Hansel, Cox, and Nugent, l98l). This work was aimed at applying three-way catalyst technology to heavy-duty gasoline engines, but some inferences regarding EGR performance in noncatalytic systems can be drawn. All testing involved a l978 model International Harvester 404-cubic-inch-displacement V-8 engine

58 Carburetor Calibration Ignition Advance Schedule EGR Typical Factory Single Exhaust System Figure ll Noncatalytic emission control (Hansel, Cox, and Nugent, l98l)

59 TABLE l8 Effect of Exhaust Gas Recirculation (EGR) On Exhaust Emissions and Fuel Consumption of 35O-Cubic-Inch-Displacement V-8 Engine with Fresh Oxidation Catalyst Emissions and Fuel Consumption, as Measured Engine Configuration on Transient Cycle (g/bhp-h) Hydro- Carbon Fuel carbons Monoxide NOX Consumption No EGR 0.47 4.0l 5.65 245 EGR No. l 0.55 5.5l 3.98 250 EGR No. 2 0.49 5.88 3.40 254 EGR No. 3 0.59 5.04 3.03 259 EGR No. 4 0.67 5.97 2.9l 254 SOURCE: Informal communication, General Motors Corporation, February 4, l98l.

60 equipped with an exhaust oxygen sensor and a prototype feedback-controlled car- buretor. In Table l9, data extracted from the same source have been arranged to show the effects of EGR on emissions in the absence of catalytic controls. The EGR system employed is the l978 production system used in the International Harvester engine, and therefore does not represent a special attempt to approach l986 standards. The results are nevertheless helpful in assessing EGR effects. Table l9 suggests that EGR can yield NOX emission reductions of 30-40 percent with no apparent compromise of hydrocarbon or carbon monoxide emissions. Fuel economy results are mixed, one set of data showing a small improvement and the other showing a small degradation with application of EGR. In both cases, fuel economy changes may be within the range of random variability of the tran- sient test procedure. It is probably reasonable to conclude that the levels of EGR incorporated have no significant effect on fuel consumption. It should be stressed in summarizing the effectiveness of noncatalytic systems that very little test data defining EGR effects on emissions and fuel consumption using the new test cycle are available. Most important, data for the high EGR rates required to approach the proposed l986 standard are not avail- able. Based on the General Motors and EPA data presented here, the committee judges that in new vehicles well-engineered systems can achieve NOX emissions of about 5 g/bhp-h, with little or no fuel economy penalty. EGR combined with engine recalibration can yield NOx levels of about 3 g/bhp-h, but this may increase fuel consumption by 3—7 percent. Catalytic Control of NOX Emissions Catalytic reduction of NOX emissions, using three-way catalyst systems similar to those used in late model passenger cars, is being considered for gasoline-powered, heavy-duty vehicles. As shown in Figure l2, these systems can take several forms, involving either single three-way catalyst beds or combi- nations of three-way catalyst beds and downstream oxidation catalysts to enhance control of hydrocarbon and carbon monoxide emissions. Data presented in this section will demonstrate that, with fresh catalysts and new engines, three-way catalyst systems are capable of approaching the NOX emission levels required by the proposed l986 standards. However, the crucial factor in compliance is catalyst durability over the engine's life as defined by the pertinent regulations. Unfortunately, so far as the committee can determine there are no meaningful data on the durability of heavy-duty three-way catalyst systems. There is strong evidence, however, to indicate that the severity of heavy-duty service may cause catalysts to deteriorate more rapidly than in pas- senger vehicle service. For this reason, emissions deterioration factors obtained from light-duty vehicles are not applicable to heavy-duty vehicles.

61 TABLE l9 Effect of Exhaust Gas Recirculation on Exhaust Emissions and Fuel Consumption of a 404-Cubic-Inch-Displacement V-8 Engine3 Emissions as Measured on Brake Specific Fuel EGR Transient Cycle (g/bhp-h) Percentage Consumption (g/bhp-h), Hydro- Carbon Reduction in With Percentage Change carbons Monoxide NOX NOX Emissions in Parentheses 350-mV Control Pointb No 4.64 43.5 8.04 322 (26.7) (-3.95) Yes 4.6l 39.8 5.89 3l0 530-mV Control Point No 4.04 49.44 7.88 304 (36.2) (+l.3) Yes 4.82 49.02 5.03 308 aEngine equipped with exhaust oxygen sensor and feedback carburetor control. "The control point refers to the setting on the electronic feedback carburetor control system, which determines the mean air/fuel ratio. (550 mV - chemically correct; 350 mV - lean.) SOURCE: Data from Hansel, Cox, and Nugent, l98l.

62 Oxygen Sensor TWC Single-Bed or Dual-Bed Catalysts Oxygen Sensor Air Oxidation Catalysts TWC Single-Bed or Dual-Bed Catalysts • To Air Meters Microswitch Oxygen Sensor TWC Catalysts Oxidation Mufflers Catalysts Figure l2 Catalytic NOX emission control systems (Hansel, Cox, and Nugent, l98l)

63 New Engine, Fresh Catalyst Emissions Data Two sets of transient test data suggest that, again with fresh catalyst, three-way catalyst systems can approach or meet the proposed l986 standards. All of the data indicate a measurable fuel consumption penalty. Table 20 represents data provided by the General Motors Corporation (l98l) for a prototype l986 system involving a 350-cubic-inch-displacement V-8 engine equipped with dual three-way catalytic converters and an EGR system for comparison purposes, data from a l984 prototype system using dual oxidizing converters are also tabulated. With the three-way system, NOX emissions were less than 2 g/bhp-h, while hydrocarbon and carbon monoxide levels were maintained below l984 limits. Fuel consumption increased by about 5 percent compared with that of the l984 prototype with no EGR, but was comparable to that of the l984 engine when EGR was added to bring NOX emissions to about 3 g/bhp-h. It should be stressed that these results reflect the performance of fresh catalyst materials with less than l0 hours of aging. Data obtained by the EPA (Hansel, Cox, and Nugent, l98l) for a prototype three-way catalyst system are shown in Table 2l. These tests employed an Inter- national Harvester 404-cubic-inch-displacement V-8 engine equipped with an exhaust oxygen sensor and experimental feedback-controlled carburetor. Both single three- way catalyst systems and dual-bed systems combining three-way and oxidation catalysts were tested. The air/fuel ratio was controlled electronically and is characterized in the table by a "millivolt control point." Lower values represent leaner mixtures, and higher values richer mixtures. Stoichiometric conditions are represented by a value of about 550 mV. To obtain adequate maximum power output, the carburetor control loop was opened at wide-open-throttle conditions to enrich the mixture. Such enrichment, however, raises oxidation catalyst temperatures and could limit catalyst life. EGR was provided by the l978 production system, which was deactivated for some tests and activated for others. All data in Table 2l reflect tests with EGR. The data in Table 2l show that, with appropriate control of the air/fuel ratio (500-700 mV), the three-way catalytic converter can yield NOX emissions of l-2 g/bhp-h with fresh catalyst. However, for this particular installation, with air-fuel ratio adjustments in the indicated range, it was necessary to add a downstream oxidizing catalyst bed to control emissions of hydrocarbons and carbon monoxide. Fuel economy for the three-way catalyst tests can be compared with base engine fuel economy data, also shown in Table 2l. Operation at the richer settings, to minimize NOX emissions, appears to increase fuel consumption by as much as 3-7 percent, though apparent test-to-test variability in fuel consumption measurements makes precise determination of fuel consumption trends difficult.

64 TABLE 20 Emissions of Heavy-Duty 350-Cubic-Inch-Displacement V-8 Gasoline Engine with Catalytic NOx Control System Engine Characteristics Emissions and Fuel Consumption (g/bhp-h) l984 Prototype 350-CID V-8 with Dual Oxidizing Converters Hydro- Carbon Fuel carbons Monoxide NOX Consumption Without EGR 0.47 4.0l 5.65 245 With EGR 0.67 5.97 2.9l 254 l986 Prototype 350-CID V-8 with Three-Way Converters and EGR Full-time air 0.36 8.l4 l.97 259 Modulated air 0.54 ll.66 l.28 259 SOURCE: Informal communication, General Motors Corporation, February 4, l98l.

65 TABLE 2l Emissions and Fuel Consumption of Heavy-Duty Gasoline Engine with Three-Way Catalyst NOX Control System (Fresh Catalyst) Emissions and Fuel Consumption as Measured on Transient Cycle (g/bhp-h) Engine Configuration Hydro- Carbon Fuel Increase in carbons Monoxide NO Consumption Fuel Consumption3 l978 Production 2.58 56.0 4.89 3l2 0 350 millivoltsb Three-way catalyst with EGR 0.85 l3.9 3.l4 328 5.2 Three—way catalyst with oxidation catalyst and EGR 0.74 3.46 3.45 337 8.2 530 millivolts Three-way catalyst without EGR l.4l 22.9 l.58 320 2.6 Three-way catalyst with oxidation catalyst and EGR 0.64 3.2l l.48 32l 2.9 630 millivolts Three-way catalyst with oxidation catalyst and EGR 0.56 3.29 0.7l 335 7.3 740 millivolts Three-way catalyst with oxidation catalyst and EGR 0.68 3.60 0.74 3l9 2.3 Percentage change compared with l978 production engine. ^Carburetor set point (350 mV - lean; 550 mV = chemically correct; 630 mV - rich). SOURCE: Data from Hansel, Cox, and Nugent, 198l.

66 All of the test date of Table 2l were obtained with nearly fresh catalyst. Durability data for these systems are not available. Catalyst Durability As emphasized above, information on the durability of emission control systems is vital to assessing the feasibility of meeting given standards with catalytic converters. Unfortunately, the committee is not aware of any such data relevant to the proposed l986 heavy-duty NOX standards. Given this lack of information, it has been suggested that durability data on three-way catalyst systems in passenger cars could be applied to the heavy-duty situation. However, there is considerable data suggesting that catalysts in heavy-duty systems undergo much more severe operating conditions than those in light-duty systems. A number of factors are responsible for the relative severity of heavy-duty service, but the most important by far is the potential for excessively high catalyst bed temperatures, leading to deterioration in catalyst activity. Such concern has been voiced with regard to deterioration of the oxidation catalysts required to meet l984 hydrocarbon and carbon monoxide standards, as well as the three-way catalytic converters needed to meet the proposed l986 NOX standard. Since this report is concerned with the l986 NOX standard, this discussion will focus on three-way converters for NOX control. However, the intent is not to minimize the potential severity of the durability problems of oxidation cata- lysts. It is well known that exposure to excessive temperatures can permanently deactivate catalytic controls. For this reason, the higher exhaust system temperatures that may be encountered in heavy-duty vehicles are of major concern. Figure l3, taken from the previously cited EPA tests (Hansel, Cox, and Nugent, l98l), shows the maximum temperatures attained during the heavy-duty transient test procedure and during steady operation at maximum power (indicated by "MAP test"). Both tests subject the three-way converter to temperatures of l400- l500°F. Temperatures encountered in actual vehicle operation may be even higher than those encountered in the engine dynamometer test, depending on operating conditions, ambient temperatures, and external air flow rates. There appears to be no consensus of opinion on the maximum permissible three-way catalyst bed temperatures for acceptable durability. The referenced EPA paper states, "The maximum temperatures experienced by the [three-way catalyst] were satisfactory and within design limits." On the other hand, information submitted to this committee by the Ford Motor Company (informal communication, February 9, l98l) states that, "Temperature guide-lines for current [three-way catalysts] are l300°F sustained and momentary spikes to l450°F for rich operation. Future development efforts may permit somewhat higher temperature operation by l986." Catalyst-aging data obtained by General Motors are shown in Figures l4 and l5. In these tests, catalyst materials are aged in high-temperature furnaces with controlled atmospheres simulating the combustion products of a given stoichiometry. For fuel-lean conditions, catalyst durability is good for

67 Upstream 1 Inch Behind Front Face Downstream Upstream 1 Inch Behind Front Face Downstream Oxygen Sensor Location 15OO F 14OO°F 153O°F 146O°F TWC Cata1yst 137O°F 1575°F 145O°F 1310° F- 167O°F 1750°F- Oxidation Cata1yst 1320°F 1510°F 14OO°F Maxima During MAP Test I Maxima > During Heavy-Duty I Transient Test Figure l3 Maximum catalyst temperatures (Hansel, Cox, and Nugent, l98l)

100 1OOO° F 100 90 o 80 LLJ LU 70 60 50 AGING TIME (h) A 1OOO° F 2000° F 12 16 20 24 36 AGING TIME (h) B Figure l4 Effect of furnace aging on three-way catalyst hydrocarbon removal efficiency, after 600-second warmup: (A) in neutral atmosphere (stoichiometric + 0.l air/fuel unit); (B) in strong oxidizing atmosphere (approximately 4-5 percent O2). (Informal communication, General Motors Corporation, February 4, l98l.)

69 100 1OO0° F 12 16 AGING TIME (h) 20 24 Figure l5 Effect of furnace aging of three-way catalyst NOX removal efficiency, after 600-second warmup, in slightly rich atmosphere (stoichiometric +0.2 air/fuel unit) (Informal communication, General Motors Corporation, February 4, l98l.)

70 temperatures as high as 2000°F. However, as the air/fuel ratio is enriched to chemically correct or slightly reducing conditions, high-temperature dura- bility suffers substantially. For mixtures only 0.2 air/fuel units richer than chemically correct, catalyst activity is degraded significantly at tempera- tures as low as l500°F. Overall Assessment of Catalytic NOX Control As stated earlier, three-way catalytic converter systems can be expected to approach or meet the NOX emission levels of the proposed l986 emission standards with fresh catalyst materials. The critical question is catalyst durability in heavy-duty vehicle service. Of major concern is the catalyst's potentially severe operating temperature environment. Experimental data on the durability of three-way systems appear to be unavailable even at the laboratory engine level. While data on exhaust system temperatures are available, statements on maximum permissible bed temperatures are conflicting. Data supplied by one manufacturer, however, suggest that, for the reducing atmospheres employed in three-way catalyst systems, measured operating temperatures in heavy-duty service may exceed permissible levels. To sum up, it appears that insufficient durability data are available currently to determine the feasibility of heavy-duty gasoline engines' meeting the proposed l986 NOX standard of l.7 g/bhp-h by using three-way catalyst technology. Substantially more testing will be required, with a variety of emissions contol system configurations, before a meaningful assessment can be made. The foregoing discussion has focused on the durability of NOX reducing cataysts and has not considered a number of other development problems that must be addressed by engine manufacturers. Among these are the questions of satisfactory engine performance and durability (exhaust valve and valve seat life, octane satisfaction, maximum power output, fuel distribution, and drive- ability. In addition, manufacturers have expressed serious concerns about the durability of the oxidation catalysts used to meet l984-l986 hydrocarbon and carbon monoxide emission standards. The intent has been not to minimize the importance of these concerns but rather to focus on the question that currently appears most critical to NOX emission control in heavy-duty gasoline engines— the durability of three-way catalyst systems. CONCLUSIONS Evaluating the emission performance of new heavy-duty gasoline engines and control systems at low emission levels is made difficult by the very limited data available. For this reason, the emissions and fuel economy values cited in the following conclusions must be regarded as tentative. (All emissions data on the following are derived from the transient test procedure.)

7l NOX emissions of about 5 g/bhp-h can be obtained in new gasoline-powered heavy-duty engines with little or no fuel economy penalty. The use of EGR combined with engine recalibration can yield new-engine NOX levels of about 3 g/bhp, with fuel consumption penalties of about 3-7 percent. The higher EGR rates and retarded timing required for lower NOX levels would further increase fuel consumption and at the same time reduce engine performance. With fresh catalysts, one three-way catalytic control system has hydro- carbon, carbon monoxide, and NOX emissions of 0.54, ll.66, and l.28 g/bhp-h, respectively, with an accompanying 5-percent increase in fuel consumption relative to that of a prototype system with an NOX level of 6.5 g/bhp-h. Also with fresh catalysts, one system with both three-way catalysts and oxidation catalysts has achieved corresponding emission levels of 0.6, 3.6, and 0.7 g/bhp-h, with increases in fuel consumption of 3-7 percent. No data on the durability of catalytic NOX control systems used with heavy-duty engines are available. However, data on exhaust temperatures indicate that catalyst durability may be a significant problem. For this reason, we cannot draw conclusions about the success of these systems until durability data become available.

72 REFERENCES California Air Resources Board. l98l. "Public Hearing To Consider Amendments to Title l3, Section l956.7, California Administrative Code, Regarding Exhaust Emission Standards and Test Procedures for l984 and Subsequent Model Heavy-Duty Engines." Sacramento, Calif.: California Air Resources Board. (Staff report) January 2l. Hansel, James G., Timothy Cox, and Thomas Nugent. l98l. "The Application of a Three—Way Conversion Catalyst System to a Heavy- Duty Gasoline Engine." Warrendale, Pa.: Society of Automotive Engineers. (SAE Paper No. 8l0085.) National Research Council. l974. Report of the Committee on Motor Vehicle Emissions. Washington, D.C.: Society of Automotive Engineers. (SAE Paper No. 8l0085.) Obert, E.F. l968. Internal Combustion Engines. Scranton, Pa.: International Textbook. Patterson, D. J., and N. A. Henein. l972. Emissions From Combustion Systems and Their Control. Ann Arbor, Mich.: Ann Arbor Science.