Air Pollution, the Automobile, and Public Health (1988)

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.

' ! ' Automotive Emissions JOHN H. JOHNSON Michigan Technological University Pollution from Automobiles Problems and Solutions / 40 Emissions Standards and Control Approaches / 40 Fuel Economy / 40 In-Use Passenger Car Emissions / 41 Emissions Regulations / 42 Emission Test Procedures / 43 Emission Standards / 45 U.S. Fuel Economy Standards / 46 Vehicle and Emission Control System Technology / 46 Spark-Ignition Gasoline-Powered Vehicles / 47 Diesel-Powered Passenger Cars: Particulate Control / 48 Diesel-Powered Heavy-Duty Vehicles / 49 In-Use Vehicle and Engine Characteristics / 50 Gasoline-Powered Passenger Cars and Trucks / 50 Diesel-Powered Passenger Cars / 52 Diesel-Powered Trucks / 52 Models for Predicting Future Emissions / 53 Fuels and Fuel Additives / 54 Trends in Gasoline Fuel Properties / 54 Fuel Usage Trends / 56 Methanol-Fueled Vehicles / 56 Trends in Diesel Fuel Properties / 59 Refueling Emissions / 59 Additives / 60 Methods for Measuring the Unregulated Pollutants / 61 Sampling / 62 Analytical Methods / 63 Current Regulated and Unregulated Emissions / 64 Regulated Emissions / 64 Unregulated Emissions / 64 Summary of Research Recommendations / 70 Air Pollution, the Automobile, and Public Health. (it) 1988 by the Health Effects Institute. National Academy Press, Washington, I).C. 39

40 Automotive Emissions Pollution from Automobiles-Problems and Solutions Concern about the automobile as a source of air pollution has been expressed period- ically, but national concern was first evi- denced in the 1960s when California estab- lished the first new car emission standards. The scientific basis of this effort is the pioneering atmospheric chemistry research of A. I. Haagen-Smit, who showed that photochemical reactions among hydrocar- bons (HC) and nitrogen oxides (NOX) pro- duce the many secondary pollutants that reduce visibility and cause eye and nose irritation in the Los Angeles area. This paper reviews our current knowl- edge of automotive emi~cion~ in~l',rlin~ standards, control technology, fuel econ- omy, fuels and additives, in-use emissions, measurement methods for unregulated pol- lutants, and models for predicting future automotive emissions. Fuel economy is in- cluded because achieving high fuel econ- omy and low emissions together makes the engineering effort more difficult. Emissions and fuel economy are interrelated because both are influenced by the engine combus- tion system design. In practice, the strin- gency of emission standards determines the importance of this interrelationship. After current knowledge in each area has been reviewed, important gaps in our knowl- edge are identified and research needed to fill these gaps is described. Emissions Standards and Control Approaches MA ~ _^ _% If-· ~ Evolving emission standards have resulted in three levels of stringency, and in turn, three types of control technology. Figure 1 describes the technologies applied in each of the three phases and the general time periods in which they were applied to cars, light-duty trucks, and heavy-duty trucks (Ford Motor Co. 1985a). The percent re- duction in the HC, carbon monoxide (CO), and NOX emissions are also shown. Air/fuel (A/F) ratio, which is controlled by the carburetor or fuel injection system, is the most important variable in determin- ing emissions and in applying catalyst tech- nology. Figure 2 (Heinen 1980) is a plot of NOX, HC, and CO concentrations in the exhaust versus A/F ratio for a typical gas- oline engine. It is impossible to achieve the low emissions demanded by federal stan- dards by A/F ratio control alone since the concentrations of the three pollutants are not minimums at the same A/F ratio. In fact, when CO and HC concentrations are  a minimum, at an A/F ratio of around 16:1, NOX production is close to a maximum. Also shown is the A/F ratio for maximum power (13.5:1) and maximum fuel economy (17:1~. The region where A/F ratio exceeds 17.5:1 is the lean bum region where misfires can occur along with slow flame speeds, causing increased HC concentration. The A/F ratio effects are used in all phases of control. The stoichiometric ratio of 14.7:1 is necessary in the Phase III control using three- way catalysts since the A/F ratio must be in a narrow window within + 0.05 of the stoi- chiometric ratio to achieve high HC, CO, and NOX control efficiencies simultaneously. Exploring the lean burn region is an impor- tant area of research and development be- cause of the potential of improved fuel econ- omy and adequate emission control with only an oxidation catalyst. Fuel Economy Federal regulations also mandate automo- tive fuel economy. The period from 1968 to 1974 resulted in primary emphasis on emission control with loss of fuel economy from lower compression ratios, changes in spark timing, A/F ratio and axle ratio changes, and exhaust gas recirculation. Fig- ure 3 (Heavenrich et al. 1986) shows the U. S. fleet combined, city, and highway fuel economy data for each model year since 1974. The figure also includes infor- mation on average vehicle weight. The fuel economy from 1977 to 1980 improved al- most exactly in proportion to the decreas- ing weight of vehicles. If the data in figure 3 were normalized to the 1978 weight mix, it would show that fuel economy improve- ments leveled out in 1982. With the intro

John H. Johnson Major Emission Reduction Phases REDUCTION (%) FROM UNCONTROLLED TIME Light Heavy FRAME Car Truck Truck pre 1968 1 975 1 979 1981 1984 1 987 and beyond 41 TECHNOLOGY APPLICATIONS _ Phase I Uncontrolled to INITIAL CONTROL LEVELS HCs CO NOx · Closed crankcase controls (HCs) 72% 67% 24% · Evaporative emission controls (HCs) · Engine modifications for improved combustion (HCs and CO) · Air injection (thermactor) systems _ ,. .~ ~ ^~ Ones and ~V) · Exhaust gas recirculation (NOx) _ Phase 11 HCs CO NOx INTERMEDIATE"TECHONOLOGY-LIMITED" 86% 82% 24% LEVELS .................................. . · 1st-Phase controls _....................................... ........... .. . . · Oxidation catalysts (HO & CO control) .......................... . .. . . . ....... ................................... _ . · Early-stage electronic engine controls ................................................ ............................. . . (optimized fuel economy, driveability, .... ................. ............................. _ . . emissions control performance) ...................................... ............................. ~.................... .............................. ............. , ................................................. Phase 111 HCs CO NO ULTIMATE CONTROL LEVELS ...................................... ............ x . . . ( ar) · 1 st Phase controls ............................... ............................................. ......................... . . . . . 90% 90% 50% (Truck) · Three way catalyst systems (HCs' CO1  ................. ...... ,., ... . . ......... A . Nux) ............................. , , .............. ... .................... .. .. , ,., .............................................. .,, .................... , . ............... . . · Particulate traps (diesels) ............................. . ..................................................... ......................................................... ....... .................... , , : . . . . . , . Figure 1. Major phases in the reduction of automotive emissions. (Adapted with permission from Ford Motor Co. 1 985a. ) auction of the oxidation catalytic converter in 1975, improved fuel economy and re- duced emissions occurred simultaneously. Further emission reduction with simulta- neous fuel economy improvement contin- ues through application of new technology, especially computer engine control. In-Use Passenger Car Emissions The in-use emissions from passenger cars exceed the new car standards mandated by law. Nonetheless, emissions continue to decrease in spite of high tampering rates and fuel switching (that is, using leaded fuel in engines developed to run on unlead . Figure 5 shows EPA emission factors data as analyzed by General Motors Corp. (1985a). The measured emission concentra tions for various model years are compared to the standards that were in effect during those years. The measured NO, concentra tions follow the standards fairly well. Al though the measured HC and CO concen trations are higher than the standards, the difference between the actual emissions and the standards appears to be narrowing (al though the ratio is not decreasing) as im proved technology, more frequent inspec tion and maintenance, and better training of mechanics has occurred. Even though the overall trend of emissions is down, a ed fuel). From field surveys in 14 cities,few vehicles have high emission levels, as Greco (1985) found the overall tamperingshown in figure 6, probably because of rates and catalyst tampering rates shown in figure 4. electronic problems rather than catalyst re- moval or misfueling problems.

42 Emissions Regulations In the 1960s, motor vehicles were identified as one of the primary sources of air pollu- tants in urban areas. Emission standards for passenger cars were first imposed in Cali- - fornia in 1965. These were followed by U.S. federal standards in 1968. The 1970 Clean Air Act further imposed stringent HC, CO, and NOX reductions for 1975 and 1976. These reductions were subsequently delayed and changed by the 1974 Energy 40r ~ 35 - O 30 8 J 25 LL Lu 20 LL 11 ~ 15 Automotive Emissions Maximum power IStoichiometric mixture Maximum fuel economy Lean burn area 400C 300C Q x 200C o 1 OOC o - 800 . - 600 ~CO / _~ 400 \ \\/ -200 _ \/\ i'" ~ O ___~' I 8 10 12 lo  'A 14 16 18 20 22 AIR/FUEL RATIO c' o - 4 ~ - - Figure 2. Concentrations of HC, CO, and NOX emissions as a function of air/fuel ratio in a typical gasoline engine. (Adapted with permission from Heinen 1980, and the Society of Automotive Engineers, Inc.) and Environmental Coordination Act and the 1977 Clean Air Act Amendments. Rec- ognition of the motor vehicle as a major source of pollutants has spread to other countries, of which many have imposed diverse standards and test procedures re- flecting various degrees of stringency. The differences have come about because of dif- ferent regulatory philosophies and air quality goals, in combination with concerns about the conflicting goal of improved fuel eff~- ciency (Barnes and Donohue 1985~. Coo 4500 10L = 0 Highway 0< a` ~o Combined '~ ~ City · Weight l 75 76 77 78 79 80 81 82 83 84 85 86 MODEL YEAR Figure 3. U.S. fleet fuel economy and average vehicle weight by model year. (Adapted with permission from Heavenrich et al. 1986, and the Society of Automotive Engineers, Inc.) 4000 imp :I1 m 3500 ~ m I 3000 2500 J 2000

John H. Johnson 50 A 0- 30 a 20 llJ ~ 10 rig o Overall tampering ~ Catalyst tampering v In. 84 83 82 81 80 79 78 77 76 75 VEHICLE MODEL YEAR Figure 4. Overall and catalyst tampering rates by vehicle model year, based on 1984 survey. (Adapted from Greco 1985.) Emission Test Procedures Passenger Cars. Emissions come princi- pally from three automotive sources: the ex- haust, the fuel system (evaporative), and crank- case ventilation gases. To give the standard (maximum allowable level of emission in grams per mile) operational meaning, two ma- jor aspects must be defined: the driving cycle and the emissions sampling method. Driving cycles are discussed below and sampling meth- ods will be covered in a later section. 12 _ HO 10 _ 8 _ 6 _ 4 _ ~ 2 _ ~ O z 100 _ in 80 _ cn _ 60 _ 40 _ ~20 E u.l o ~ 5 lo: 4 3 2 _ 1 _ O ~To a-" 's80 81-82 83-8s o_ control ): VEHICLE MODEL YEAR 0 Figure 5. Average vehicle lifetime HC, CO, and NOx emissions compared with standards (STD), by model year for all industry passenger vehicles. (Adapted with permission from General Motors Corp. 1985a.) 43 Regulations require exhaust emission measurements during the operation of the vehicle (or engine) on a dynamometer dur-  ing a driving cycle that simulates vehicle road operation. The approach to driving cycles by various regulatory authorities represent two basic philosophies. Accord- ing to the first, the driving cycle is made up of a series of repetitions of a composite of various vehicle operating conditions repre- sentative of typical driving modes. The European Economic Community and Jap- anese cycles reflect this philosophy. Ac- cording to the second, the composite of driving modes is an actual simulation of a road route. The United States, Canada, Australia, Sweden, and Switzerland all use a version of the federal test procedure (FTP). The FTP cycle is divided into a "transient" portion and a "stabilized" por- tion with a total cycle time of 1,372 see, a driving distance of 7.5 miles, and an aver- age speed of 19.7 miles per hour (mph). Two such cycles are run: one with the vehicle at an ambient temperature of 1~30°C before start ("cold" cycle), and one with the engine control system hot ("hot" cycle) after a 10-min shutdown after running the cold cycle. Trucks. Many of the light-duty trucks intended primarily for the carrying of goods are also capable of use as passenger vehicles. The gross vehicle weight for light-duty trucks in the United States is less than 8,500 lb; trucks heavier than 8,500 lb 300- 200 : 100 .~_ ~ no. ... I~ CAL ~ ' 1 ' 1 ' -r ' 1 1 20,000 40,000 60,000 80,000 ODOMETER MILES Figure 6. Scatter plots of CO emissions from 703 1981 model-year federal cars. (Adapted with permis- sion from General Motors Corp. 1985a.)

44 Automotive Emissions Table 1. Motor Vehicle Emission Standards in the United States Emission Ratesa Federal California Evap Partic- Evap Partic Model HC CO NOX (g/ ulatesb HC+ HC CO (g/ NO.` (g/ ulatesb HC+ Year (g/mi) (g/mi) (g/mi) test) (g/mi) NOy (g/mi) mi) (g/mi) test) (g/mi) NOr Passenger Cars Pre control 10.60C 84.0 4.1 47 10.60' 84.0 4.1 47 1966 6.30 51.0 (6.0) 1968 6.30 51.0 (6.0)d 6.30 51.0 1970 4.10 34.0 4.10 34.0 6 1971 4.10 34.0 4.10 34.0 4.0 6 1972 3.00 28.0 2.90 34.0 3.0 2 1973 3.00 28.0 3.0 2.90 34.0 3.0 2 1974 3.00 28.0 3.0 2.90 34.0 2.0 2 1975 1.50 15.0 3.1e 2 0.90 9.0 2.0 2 1977 1.50 15.0 2.0 2 0.41 9.0 1.5 2 1978 1.50 15.0 2.0 be 0.41 9.0 1.5 6 1980 0.41 7.0 2.0 6 0.39f 9.0 1.0 2 1981 0.41 3.4 1.0 2 0.39 7.0 0.7 2 1983 0.41 3.4 1.0 2 0.398 7.0 0.4 2 1984 0.41 3.4 1.0 2 0.39 7.0 0.4 2 0.60 1985 0.41 3.4 1.0 2 0.39 7.0 0.4 2 0.40 1986 0.41 3.4 1.0 2 0.60 0.39 7.0 0.4 2 0.20 1987 0.41 3.4 1.0 2 0.20 0.39 7.0 0.4 2 0.20 1989 0.41 3.4 1.0 2 0.20 0.39i 7.0 0.4 2 0.08 Light-Duty Trucks 1975 2.00 20.0 3.1 2 2.0 1976 2.00 20.0 3.1 2 0.90 17.0 2.0 1978 2.00 20.0 3.1 6 0.90 17.0 2.0 6 1979 1.70 18.0 2.3 6 0.50j 9.0 2.0' 6 1980 1.70 18.0 2.3 6 0.50/ 9.0 2.0' 2 1981 1.70 18.0 2.3 2 0.50 9.0 1.5 2 1983 1.70 18.0 2.3 2 0.50 ~9.0 1.0P 2 1984 0.80 10.0 2.3 2 0.50 9.0 1.0P 2 1986q 0.80 10.0 2.3 2 0.50 9.0 1.0 2 1988 0.80 10.0 1.7 2 0.50 9.0 1.0 2 (Table continued next page.) NOTE: Evap = evaporative HC. a Emission rates for HC, CO, NOB by (or adjusted to equivalent) 1975 Federal Test Procedure. b Diesel passenger cars only. c Crankcase emissions of 4.1 g/mi not included; fully controlled. NOX emissions (no standard) increased with control of HCs and CO. e Change in test procedure. f NMHC standard (or 0.41 g/mi for total HC). g 0.7 NO.~ optional standard 1983 and later, but requires limited recall authority for 7 yr/70,()00 mi. h Optional standard, 0.3 g/mi, requires 7 yr/75,000 mi limited recall authority. ' Primary standard = 0.4 g/mi required on 90% of production after 1989. 0.41 for <4,000 lb. k 1.5 for <4,000 lb. 0.39 for <4,000 lb (nonmethane) and in following years. m 0.6 for >6,000 lb and in following years. n 2.0 for >6,000 lb. ° California: 1.0 NOR optional standard for 1983 and later, but requires limited recall authority for7 yr/75,000 mi. Primary standard = 0.4 g/mi.

John H. Johnson 45 Table 1. Continued Emission Rates Federal California Partic- Partic HC CO NOX ulates HC ulates (g/ (al (g/ Evap (g/ (g/ CO NO. ~Evap (g/ Model bhp- bhp- bhp- (g/ bhp- HC+ bhp- (g/bhp- (g/bhp- (g/ bhp- HC+ Year hr) hr) hr) test) hr) NOy hr) hr) hr) test) hr) NO. Heavy-Duty Truck and Bus Engines' 1969 " '' 1972 . v 1973 40.0' 16.0 1974 40.0 16.0 40.0 16.0 1975 40.0 16.0 30.0 10.0 1977 40.0 16.0 1.0 25.0 7.5 5.0 1978 40.0 16.0 1.0 25.0 7.5 6 5.0" 1979 25.0 10.0 1.5 25.() 7.5 6 5.0 1980 1.5 25.0 10.0 1.0 25.0 7.5 2 6.0 1984 1.5 25.0 10.0 0.5 25.0 7.5 2 4.5"' 1985 1.9 37.1- 10.6Z 3cia 0 5 25.0 7.5 2 4 5i''bb 1987 1.1 14.4 10.6 3 0.5 25.0 7.5 2 4.5 1988 1.1 14.4 6.0 3 0.6 0.5 25.0 7.5 2 4.5 1991 1.1 14.4 5.0 3 0.25 0.5" 25.0" 7.5" 2 4.5 NOTE: Evap = evaporative HC. P 1.5 for >6,000 lb. q Full useful life requirement = 11 yr/120, 000 mi (was 5 yr/50,000 ml). r NOX federal standard = 1.2 g/mi under 3,751-lb loaded vehicle weight (LVW), 1.7 g/mi for 23,751 lb L~W, and 2.3 g/mi for '6,000 lb LVW. s 1.2 for <3,751 lb. ' Various test methods, values are not strictly comparable. U275ppmHC, 1.5% CO. v 180 ppm HC, 1.0% CO. w A combined standard is optional in lieu of separate HC and NO.~ standards (for example, 1 g HC + 7.5 g NO~ or 5 g tHC+NOxi). x 1.3 for diesel. Y 15.5 for diesel. Z 10.7 for diesel. aa 4.0 for >20,000 lb. bb Gasoline only and in following years. cc 1988 federal standards for NO y have been postponed until 1990. Separate standard of 0.1 for all 1991 urban buses and all 1994 engines. ee 1.3 HC, 16.5 CO, 5.1 NO~ for diesel. SOURCE: Adapted with permission from General Motors Coro. 1986. are classified as heavy-duty vehicles. The driving-cycle philosophies for the light commercial vehicles follow those for pas- senger cars. For heavy commercial vehi- cles, engine dynamometers are used, not chassis dynamometers; that is, the enp;ine rather than the vehicle is certified. The new (effective 1985) U.S. transient test proce- dure for heavy-duty vehicles combines the two philosophies just described in that the cycle is made up in a random way from actual driving cycle data. The use of this cycle replaces the 13-mode steady-state cy- cle in use since 1973 in California and since 1974 nationally (U.S. Environmental Pro- tection Agency 1972~.  Emission Standards United States. Emissions standards and test procedures in the United States have changed significantly since the first auto- mobile emission standards were imposed in California in 1966 (see table 1) (General

46 Automotive Emissions Motors Corp. 1986~. Light-duty truck stan- dards are somewhat higher than the car stan- dards because of the differences in weight. The U.S. passenger car regulations re- quire that the vehicle comply with the emission standards for five years or 50,000 miles, whichever occurs first. Certification testing of prototype vehicles for 50,000 miles of use is based on the Automobile Manufacturers' Association (AMA) 40.7- mile durability cycle. The cycle consists of numerous stops, acceleration, and high/ medium-speed driving (maximum of 55 mph) (U. S. Environmental Protection Agency 1973~. Europe. The European Economic Com- munity, an inter-Europe regulatory body, has announced future model standards for passenger cars based on three engine size (displacement) categories. Large-car (> 2- liter engine displacement) standards are roughly equivalent to current U.S. stan- dards although there is no valid correlation between the distinct U.S. and European emission test cycles. Standards for medium cars (1.~2.0 liters) are considered to fall in the Phase I/Phase II range shown in figure 1. Requirements for small-car levels (< 1.4 liters) are comparable to Phase I require- ments. The standards include diesels; how- ever, large diesel cars are only required to meet medium-car levels. Japan. Catalyst forcing standards cur- rently in effect for passenger cars are 0.25 HC/2.1 CO/0.25 NOX g/km for the unique 10-mode hot start and 7.0 HC/60 CO/4.4 NOX g/test for the 11-mode cold- start test procedures. These standards are generally considered to be equivalent to current U.S. California levels (Ford Motor Co. 1985a). U.S. Fuel Economy Standards There have been passenger car and light- truck fuel economy standards since 1978 and 1979, respectively. The manufacturers are required to conduct passenger car fuel economy tests according to the U.S. Envi- ronmental Protection Agency (EPA) urban or "city" driving cycle the FTP for emis sion testing described earlier. The EPA also has a suburban or "highway" cycle that includes a significant amount of simulated highway driving. A combined fuel econ- omy number based on these two tests is published by the EPA and the U. S. Depart- ment of Energy and used by manufacturers in their sales literature. Manufacturers each have to meet the Corporate Average Fuel Economy (CAFE) standards for their sales-weighted fleet. Car  standards started at 18 miles per gallon (mpg) in 1978, went to 27.5 mpg in 1985, but were reduced to 26 mpg by the U. S. De- partment of Transportation for 198~1988. Vehicle and Emission Control System Technology The technology used for emission control in cars changed rapidly in the 1970s as the automotive industry spent considerable re- search and development funds to meet the stringent emission standards originally set by the 1970 amend 1977 Clean Air Act Amendments. This technology is now be- ing optimized to reduce the product cost associated with emission controls while im- proving the in-use durability of the emis- sion control systems. Heavy-duty gasoline- powered vehicles have used this technology as allowable emissions have progressively decreased. Control technology is being developed to meet proposed standards and anticipated changes in fuels. Proposed 1988, 1991, and 1994 particulate standards require new con- trol systems for heavy-duty diesels. For the United States to become less dependent on imported petroleum fuels, there is interest in using methanol in passenger cars and diesel-fueled buses. There are continued efforts to develop stratified-charge engines for passenger cars because of their potential for better fuel economy at equivalent emis- sions. There is also a demand for develop- ment of direct-injection diesels that give 15 percent better fuel economy than pre- chamber or swirl-chamber engines with equivalent or better emissions. An addi- tional demand exists for an adiabatic diesel

John H. Johnson 47 engine (more precisely, a low-heat-re- jection engine) that would have improved fuel economy and lower emissions with a simpler cooling system, particularly for vehicles in the heavy-duty class. Spark-Ignition Gasoline-Powered Vehicles During the past 15 years, emissions have been significantly lowered by improved design of the engine and fuel system while still achieving the high fuel economy de- manded by the federal standards and the consumer market. These reductions have come about by A/F ratio control, cylinder- to-cylinder distribution of air and fuels choke operation, combustion chamber de- sign, fuel injection, exhaust gas recircula- tion (EGR), ignition systems, spark tim- ing, valve timing, and many additional design details. The computer scheduling of spark timing, EGR, A/F ratio, and trans . . . . . . mission gear ratio as a tunctlon ot engine operating conditions are done very pre- cisely with sensors and actuators. This scheduling is referred to as the engine cali- bration. With all of this technology, ve- hicles still do not meet HC/CO/NOX standards of 0.4/3.4/1 g/mi without after- treatment devices. The period after 1983 has seen better optimization of systems and removing of components to reduce costs, but nevertheless, catalysts are still neces sary. Catalyst Control Systems. Meeting the 1975 MC/CO standards of 1.5/15 g/mi and at the same time increasing the fuel econ- omy was achieved through the broad intro- duction of the oxidizing catalytic con- verter. The catalyst is cold (1~30°C) at the start of the FTP cycle and must warm up to 250-300°C before oxidation of CO and HC occurs. The time required for this is a function of catalyst design and position but can be from 20 to 120 sec. The HC emitted during this period can be one-fourth to three-fourths of the allowable limit (Hil- liard and Springer 1984~. The amount of NOX emitted during the cold start is only about 10 percent of the allowable limit. The time period from 1975 to 1984 saw ~ 00 ~ ' NOx ,, CO z t) 60 _ IL z an _ o En cr o  80 40 20 _ ~ Window O , ~ , // I I 13 14 15 16 AIR/FUEL RATIO Figure 7. Conversion efficiency characteristics of a three-way catalyst. (Adapted with permission from Amann 1985.) increased fuel economy and improved emission control through exploitation of the high HC and CO removal efficiency of the oxidizing catalytic converter, so that the engine calibration could be optimized for efficiency. Progress was made by de- creasing the cold-start engine-out HC and CO emissions, by achieving faster con- verter light-off, by reducing heat loss from the exhaust system, and by reducing the deterioration of catalyst performance with cumulative driving distance (Amann 1985~. Reducing combustion temperature by spark retard and/or diluting the incoming mixture with EGR provided NOX control during the time period from 1973 to 1980. The 1981 standards stipulate no more than 1 g/mi NOX, which could not be achieved either by EGR or engine design and cali- bration. Two additional catalytic ap- proaches have gained widespread applica- tion along with the microprocessor control system, to provide the necessary control: the "three-way" catalyst and the "dual" catalyst. Three-way catalysts are capable, within a narrow range of exhaust stoichiometry, of simultaneously decreasing NOX, HC, and CO, as shown in figure 7. Within a narrow range of values of the A/F (approximately + 0.05 from the optimum), all three emis- sions are decreased with a reasonably high efficiency. An oxygen sensor is used in the . . . . . . ex Bust in conjunction Wit ~ a mlcroproces sor to make this technology feasible.

48 Automotive Emissions CARBURETOR Al R FU EL DRIVER ~ AIR 1 l ETERIN l:' ELECTRONIC SIGNAL ITHROTTLE I DEVICE I 1 CONTRA LEA ENGINE mu\ 1 \ - V ~ OXYG EN SENSOR Figure 8. System for closed-loop control of A/F ratio. The oxygen sensor inserted in the exhaust pipe ahead of the catalyst measures oxygen concentration and signals the electronic controller to adjust fuel rate continuously. (Adapted with permission from Amann 1985.) In a dual catalyst, two catalysts are used in series- a three-way catalyst followed by an oxidizing catalyst. Air is injected into the exhaust gas between the two catalysts to provide the oxygen necessary for the oxidizing catalyst to operate efficiently. Once more, precise A/F ratio control is required to make the three-way catalyst function. During the cold-start portion of the FTP cycle, the air supply to the oxidiz- ing catalyst can be diverted to the exhaust ports to add oxygen to the combustion products of the rich start-up mixture for faster catalyst light-off and to achieve higher HC and CO control efficiencies in the three-way catalyst. The dual-bed con- verter is more complex than the single-bed three-way catalyst, because it requires an . . extensive air management system. To maintain A/F ratio control within the narrow window, closed-loop control (feed- back control of fuel delivery on oxygen level in exhaust) was introduced on many cars in 1981. The schematic of a typical system is shown in figure 8 (Amann 1985~. The key element in the closed-loop system is the oxygen sensor inserted in the exhaust pipe ahead of the catalyst. It measures exhaust oxygen concentration and signals an electronic controller to adjust fuel rate continuously so that the mixture is main- tained at the stoichiometric ratio. Current Control Approaches. Since 1983 the number of engines with some type ot No particular trend in emission systems is evident except for the use of heated oxygen sensors to initiate closed-loop operation faster and more predictably and to maintain it during long idling periods. The heated sensors also deteriorate less with extended mileage (Way 1985~. Most cars use closed loop control with a three-way catalyst;  many also have an oxidation catalyst that is a dual catalyst and one of three air supply systems (pulse air, air pump, or pro grammed pump). Lean-Burn Combustion Systems. An im portant engine emission control system un der development is the lean combustion system. This system uses a closed-loop microprocessor in conjunction with lean mixture sensor and an oxidation catalyst. This alternate emission control approach achieves good fuel economy (potential 10-15 percent improvement) and also meets the emission standards by operating beyond 22:1 A/F where NOX emission is low enough to meet the 1-g/mi standard. In this lean operating region, the engine needs a different sensor design to provide feedback, and also a highly turbulent fast burn combustion system so that slow flame speed and misfires do not cause emissions and driveability problems. Toyota has de veloped and marketed such a system in Japan but not yet in the United States (Kimbara et al. 1985~. It may be possible to introduce this type of system into the U. S. market, but dura bility and driveability under hot and cold conditions need to be examined further (Kimbara et al. 1985~. The other important technological limit might be that lean burn could be restricted to cars under 2,500 3,000 lb because NOX generally increases with vehicle weight. Diesel-Powered Passenger Cars: Particulate Control There has been a major research and devel opment effort during the past seven years to develop aftertreatment devices for diesel passenger cars to meet the federal 0.2 g/mi tuel Injection has grown drastically, but standard first proposed for 1985 and later carburetors are still used on many engines. put offuntil 1987. California has a 0.2-g/mi

John H. Johnson r G'owp~ug Controller ~ driver to Battery _~ RPM ~ Power ~ control Pressure box ') to Glowplug driver / ~signal '\bielow) Fox. ,~ ~ Pressure /__~/' \ transducer Glowplug igniter (air Figure 9. Diesel particulate trapping system utiliz- ing a ceramic fiber trap, a fuel additive, glow plug igniters, and exhaust backpressure regeneration con- trols. (Adapted with permission from Simon and Stark 1985, and the Society of Automotive Engineers, Inc.) standard that was initiated in 1986, and will be lowered 0.08 g/mi in 1989. A number of prototype systems have been built and field tested to meet the 0.2 g/mi standard. Mer- cedes-Benz (Abtoff et al. 1985) introduced a catalytic trap oxidizer in 1985, in conjunc- tion with careful modification of the engine (in particular, the turbocharger). The sys- tem meets and is certified to the 1986 49 California standards and has been sold in the 11 western states. Volkswagen has de- veloped a prototype system that uses a Corning ceramic particulate filter in con- junction with Lubrizol 8220 manganese  (Mn) additive. The additive consists of nonstoichiometric Mn fatty acid salts dis- solved in naphtha, which is metered from a separate fuel-additive storage tank on the vehicle (lifetime filling) and mixed with the fuel (Wiedemann and Neumann 1985~. Emissions of Mn oxide of all valence states, as well as MnSO4, may occur. Data sug- gest that most of the Mn residue is in the form of sulfate. General Motors has also tested a system, shown in figure 9, with on-board tank- blending, additive dispensing, and ceramic fiber trap (Simon and Stark 1985~. This system uses pressure and engine speed to provide a measure of particulate loading for triggering the glow plug igniters for regen- eration. Simon and Stark (1985) investi- gated three different additives: cerium (0.13 g Ce/liter), manganese (0.07 g Mn/liter), and cerium plus manganese (0.07 g (Ce + Nln)/liter). Their tests showed that vehicles equipped with properly tuned 4.3-liter en- gines and operated using a fuel additive would not, on a production basis, be able to meet the 1987 federal emissions stan- dards at sea level or at altitude. Equipped with particulate traps, however, the vehi- cles would probably meet the 1987 federal standards and might, with further engine tailoring, be able to meet the 1989 Califor- nia standards on a production basis. Diesel-Powered Heavy-Duty Vehicles Diesel-powered heavy-duty vehicles use direct-injection turbocharged engines of two-cycle as well as four-cycle design. Die- sel engines are designed for a commercial market and hence durability, reliability, and fuel economy drive their development. The approaches enforced to date to meet the standards for particulates, HCs, and NOX have involved improved turbocharg- ers, intercooling, improved fuel systems and nozzles, and electronic fuel injection control. To reduce NOX emissions and improve fuel economy, some manufactur

50 Automotive Emissions ers use heat exchangers to lower air inlet temperature. The industry believes that the 1988 standards can be met with advanced electronic fuel systems and possibly with mechanical fuel systems with electronic- governing, air-to-air intercoolers (or low- flow radiators) and improved turbocharg- ers, but that the 1991 particulate standard of 0.25 g/brake horsepower (bhp)-hr will need trap technology. EPA standards have emphasized particulate control rather than NOx control. There is the feeling that the 1988 standards will result in some loss of fuel economy. GM-Detroit Diesel Allison Division (DDAD) has also decided to re- move their 2-cycle engines from the on- highway market because of the disadvan- tages of this engine under these tight emissions constraints. This is now being reevaluated under the new Detroit Diesel Corp. In-Use Vehicle and Engine Characteristics How vehicles and engines perform in the hands of the operator ultimately determines their emissions and, in turn, their impact on air quality. This section examines the emission characteristics of gasoline- and diesel-powered passenger cars and trucks as actually used by owners. The effects of field environmental conditions such as tempera- ture, tampering (removal of and changes in components), or misfueling on emissions are discussed. Gasoline-Powered Passenger Cars and Trucks To develop an understanding of the in-use characteristics of gasoline-powered passen- ger cars, it is important to know whether tampering and misfueling with leaded fuel occur. Misfueling has a twofold impact on the environment: increased lead (Pb) emis- sions and increased regulated HC, CO, and NOx emissions due to poisoning of the catalyst. Tampering has a direct effect on the regulated emissions and can also affect the unregulated emissions. In this section, Table 2. Tampering Prevalence in Light Duty Vehicles for Critical Emission Control Components, April-October 1984 Component/System Tampering Rate (%) Trucks Cars Overall Catalytic converter Filler neck restrictor Air system PCV system Evaporative control system  EGR system 10 10 10 Overall 14 14 12 3 5 2 9 10 7 2 2 27 21 22 Fuel switching 19 13 14 SOURCE: Adapted from Greco 1985. the latest tampering data gathered by the EPA and the Motor Vehicle Manufactur- ers' Association (MVMA) are examined first. This is followed by data showing the emissions and fuel economy of vehicles in use. Tampering and Misfueling. The latest EPA report on tampering is based on a survey of 4,426 light-duty vehicles con- ducted in 14 cities between April and Oc- tober 1984 (Greco 1985~. These inspections were performed with the consent of the vehicle owners and therefore may underes- timate tampering rates. Four categories were used to summarize the condition of the inspected vehicles: 1. Tampered at least one control device removed or rendered inoperative; 2. Arguably tampered possible but not clear-cut tampering; 3. Malfunctioning; 4. Okay all control devices present and apparently operating properly. Greco's overall survey averages of vehi- cle condition were as follows: tampered, 22 percent; arguably tampered, 29 percent; malfunctioning, 4 percent; okay, 46 per- cent. The rates for tampering with selected components and the rates of fuel switching are shown in table 2. These results have not been weighted to compensate for inspec . . lion anc maintenance program representa tion and probably underestimate the actual

John H. Johnson 51 Table 3. Incidence of Misfueling in Large Urbana and Nonurban Areas, 1981-1982 Overall Catalyst-Equipped Fleet Occasional Misfuelers Persistent Misfuelers Purchase Vehicle Vehicle Vehicle Area Volume Involvement Leaded Fuel Involvement Rate (%) Rate (%) Usedb (%) Rate (%) Leaded Fuel Involvement Leaded Fuel Used (%) Rate (%) Used (%) Large Urban 4.0 14.0 21.6 7.3 3.7 1.6 9.6 Nonurban 11.8 22. 1 41.8 5.9 1.9 6.7 28.1 Total U.S. 7.5 18.0 100.0 7.1 8.6 3.5 53.1 a Large urban areas are defined as standard metropolitan statistical areas with populations over 1 million. In the National Panel Diary Inc. data base these areas account for about one-half the total urban vehicle population. b Percent of all misfueled leaded fuel purchased by the entire catalyst-equipped fleet. SOURCE: Adapted with permission from McNutt et al. 1984, and the Society of Automotive Engineers, Inc. nationwide rates. The tampering rates for catalytic converters and filler inlet restric- tors (the insert in the fuel tank neck that prevents insertion of the larger leaded fuel nozzle) have increased steadily since 1978, whereas the rates for other components have fluctuated. The increasing tampering rates for catalytic converters and inlet re- strictors may be partly due to the increasing age of the vehicles surveyed. In addition, the presence of inspection and maintenance programs affected tampering rates. The catalyst was removed in 3 percent of the vehicles in areas with mandatory inspection and maintenance programs and in 11 per- cent of the vehicles in areas having no programs. Removing the catalytic converter in- creases HC and CO emissions by an aver- age of 475 percent and 425 percent, respec- tively (U. S. Environmental Protection Agency 1983~. For vehicles equipped with three-way catalysts, substantial increases in NOx emissions would also be expected to occur. Tampering with the EGR system can increase NOx emissions by an average of 175 percent (Greco 1985~. Fuel switching, defined as the presence of a tampered fuel filter inlet restrictor, a positive Plumbtesmo tailpipe test, or a gas- oline Pb concentration of more than 0.05 g/gal, was found in 14 percent of the unleaded gasoline-powered vehicles in the 1984 survey (see table 2~. Regional distri- bution in the prevalence of misfueling is shown in table 3. The impact of fuel switching on emissions depends upon its duration and certain vehicle characteristics, but emission increases of 475 percent for HCs and 425 percent for CO can easily occur (Greco 1985~. The tampering rate for light-duty trucks was equal to or higher than that for auto- mobiles in every tampering category, as  shown in table 2. The difference in preva- lence of catalytic converter tampering is particularly striking nearly three times as prevalent in light-duty trucks as in passen- ger cars (14 percent versus 5 percent) (Greco 1985~. To confirm the EPA tampering and mis- fueling data, the MVMA recently studied catalyst removal and defeat of the fuel filler restrictor. The vehicles used in the MVMA survey were a sample of 1975-84 model year cars and light-duty trucks from scrap- yards and impoundment areas in 10 cities (Motor Vehicles Manufacturers' Associa- tion 1985; Survey Data Research 1985~. The MVMA study sampled 1,865 vehi- cles, allowing the following conclusions to be reached to a 95 percent confidence level by Survey Data Research (1985~: 1. Nationwide, 8.3% of all the vehicles in the sample were found to have their catalytic converters removed. This removal rate is signif- icantly higher among older (i.e., 1975-1978 model year passenger cars and light-duty trucks. 2. The rate offuel filler neck restrictor tam- pering on a national basis /~7.3%) is slightly lower than the rate of catalytic converter removal (8.3%J. Again, this tampering rate is higher among older (i. e., 1975-1979) model year cars and li~ht-duty trucks. 3. O , Both catalytic converter removal andfuel filler neck restrictor tampering rates are substan- tially lower in the sample of Inspection/Mainte

52 Automotive Emissions nance area locations than the sample of Non- Inspection/Maintenance area locations. As a result of this study, MVMA is now confident that the much more detailed EPA studies, covering in addition such compo- nents as the air pump, EGR system, the positive crankcase ventilation (PCV) sys- tem, the evaporative emissions control sys- tem, and others, are yielding results that are reasonably representative of the in-use fleet (Motor Vehicles Manufacturers' Associa- tion 1985~. Elects of Tampering on Emissions. Re- cently, the Automobile Club of Southern California conducted a test program using its 1981 fleet vehicles (General Motors, Buick, and Pontiac vehicles) in an effort to better understand the effect of system com- ponent failures. The primary objective of the program was to determine the degree to which fuel economy, exhaust emissions, horsepower, and driveability are affected by disabling key components of a comput- er-controlled system; a secondary objective was to establish a method of accurately and efficiently identifying vehicles with dis- abled components Jones et al. 1982~. Jones and coworkers (1982) found that disabling the coolant temperature sensor, the throttle position sensor, or the mixture control solenoid has a major effect on ve- hicle performance. Disconnection of the coolant temperature sensor increased HC emissions an average of 549 percent and CO emissions an average of 1,120 percent over baseline emission levels; disconnection of the throttle position sensor increased HCs by 1,195 percent and CO by 3,113 percent; and disconnection of the mixture control solenoid increased HCs by 1,293 percent and CO by 3,438 percent. Each of these disablings is the disconnection of a single electrical connector Jones et al. 1982~. · Recommendation 1. Tampering and Misfueling. Tampering and misfueling sta- tistics are fairly well developed but their effect on emissions is not as well known. Therefore, work should be done to better characterize the effect of tampering and misfueling on emissions from vehicles and to better assess their effect on ambient pollutant concentrations. Diesel-Powered Passenger Cars The available in-use data are much more limited for diesel passenger cars than for gasoline-powered cars for two reasons: first, there are far fewer of them, and second, most of the diesel cars that are in  use were sold between 1979 and 1983 so only a small proportion of them are more than seven years old. Hyde and coworkers (1982) drew the following conclusions about the relation between cumulative mileage and rate of emissions from a sample of 20 in-use light- duty diesel vehicles from General Motors, Volkswagen, and Mercedes-Benz. 1. Particulate emissions do not show a mileage-related deterioration (increase) in the Volkswagen group and the Mercedes- Benz group, but show a large deterioration in the General Motors group because of a . . . . arge increase In extract emissions. 2. Federal Test Procedure HC emissions do not show a mileage-related deterioration in the Volkswagen and Mercedes-Benz groups, but show a deterioration in the General Motors group. 3. FTP CO emissions show a deteriora- tion in the General Motors and Volkswagen groups but not in the Mercedes-Benz group. 4. FTP NO',. emissions show a decrease with accumulated mileage in the General Motors and Mercedes-Benz groups but no change for the Volkswagen group. Diesel-Powered Trucks There are limited data on diesel engines in operational use although there are some recent laboratory data obtained by the var- ious manufacturers in an EPA/Engine Manufacturers Association (EMA) in-use emission factor test program for heavy- duty diesels. As part of this emission factor testing effort, the following classifications were included in the sample of 30 engines: tampered engines, poorly maintained en- gines, and rebuilt engines. The engines

John H. Johnson - ~ ~ ~ LE2 passenger car \ MVMA MOBILE1 \ \ ~pas,~r car 10: o 0 20 at o c' C, 30 G 40 _ rEPA MOBILE1 passenger car 16 highest U.S. stations air quality - \ 50 , , , , . 1 1 1973 1974 1975 1976 1977 1978 1979 .1980 YEAR Figure 10. CO air quality and emission factor trend as calculated from three computer models and as measured from the base year 1973. CO measurements were averaged from 50 U.S. stations and from 16 U.S. stations reporting the highest 8-fur yearly concentrations. (Adapted from General Motors Corp. 1985a.) were tested on FTP 13-mode steady-state and FTP Heavy Duty transient cycle. The EMA reached the following tenta- tive conclusions from this program (Gen- eral Motors Corp. 1985b): 1. The in-use control of gaseous emis- sions from heavy-duty diesel engine from the 1979-80 model year is quite good. 2. Tampering and poor maintenance do not result in excessive gaseous emissions. 3. The lab-to-lab variability of transient emission test results of unburnt HCs as well as particulates needs to be improved. ~ Recommendation 2. Diesel Particu- late Emissions and Control. There is a need for continued research on particulate con- trol technology, including the regeneration systems, to reduce the cost and complexity of these systems and the associated fuel economy penalties. Work needs to con- tinue with various additives, substrate ma- terials, regeneration systems, and controls to develop optimum systems that are able to decrease the diesel particulate emissions to the levels of 0.1 g/bhp-hr for heavy-duty  diesels and 0.08 g/mi for light-duty vehi- cles required in California. In conjunction with this research there is a need to measure 53 the metal species and the size distribution of the particles coming from diesel particulate traps. Models for Predicting Future Emissions Computer models are used for predicting future emissions from in-use vehicles. The EPA publishes the vehicle emissions model most used at present (U.S. Environmental Protection Agency 1985~. The highway source data are based on MOBILES, a computer program issued by the EPA in June 1984 and recently updated (U.S. En- vironmental Protection Agency 1985~. Fig- ure 10 shows the predicted trends of vari- ous models compared to actual air quality data for CO. The curve for percent reduc- tion of CO predicted by MOBILE2 does not correspond to the curves generated by air quality measurements from 50 U. S. stations or the 16 highest U. S. stations (General Motors Corp. 1985a). There are large differences between Gen- eral Motors' analysis of the actual emissions data and the EPA MOBILES emission factor data beyond 50,000 miles, with

54 Automotive Emissions MOBILE3 being the higher (General Mo- tors Corp. 1985a). General Motors attri- butes the difference to EPA's choosing too high a bhp-hr/mi constant for gasoline-as well as diesel-powered heavy-duty vehicles. In addition, General Motors is concerned that the evaporative submodel in MOBILE3 uses a value of 11.5 psi for Reid Vapor Pressure (RVP) for gasoline volatility whereas the national average is 10.5 psi. In fact, the RVP varies with the season of the year in different parts of the country as formulations matched to seasonal condi- tions are refined and delivered to the pumps. It is unlikely that the model will ever give good results if a single RVP number is used to represent evaporation characteristics in all places at all seasons. Instead, the United States should be subdi- vided into the American Society for Test- ing Materials (ASTM) class regions to al- low for seasonal changes in RVP. The model would then use RVP values that are representative of the season and region of the country. Other problems include esti- mating the number of trips per day for an average vehicle, identifying an appropriate ambient temperature, and understanding the effect of fuel aging. Furthermore, a new approach using a proportion of vehicles in each model year with emission rates in each of a number of incremental ranges, that is, a distribution for emission rates within each model year, should be developed for modeling emission rates. The model needs to account for the few high-emission vehicles as well. The evaporative emissions submodel needs additional work so that it better simulates the actual field fuel and control system effects, because the actual and test fuels have different vapor pressures. Re- gional and seasonal differences in RVP should be incorporated in the model along with the effects of alcohols. ~ Bon 3. Evaporative Emis- sion Model. An improved vehicle evapora- tive emissions model should be developed that is valid over various types of operating conditions for a variety of ambient tempera- tures. At the same time, changes should be made in the EPA test procedure to obtain the Average of Class C cities 0 Average of Class B cities - 13 A _ - cn 12 0 Average of Class A cities  a, cr) Cat 1 1 _ ~ 10 m IIJ tar 9 On cn ~ 8 _ cr: o CL ~ 7 _ +2a limits -. ~ / o~'' 1976 1978 1980 1982 1984 Class C ' limit .. Class B limit Class A limit YEAR Figure 11. Trends of gasoline RVP averaged by class of cities. Classification of cities by the ASTM D439 is based on weather conditions and geographical location. (Adapted with permission from Ford Motor Co. 1985b.) data necessary to properly design and size the evaporative system for the high-temperature soak situation, and data should be sought that can be used in EPA's MOBILE3 computer model for other use patterns of cars. Fuels and Fuel Additives Trends in Gasoline Fuel Properties The EPA limited the use of Pb in gasoline to 0.5 g/gal after July 1, 1985, and to 0.1 g/gal after January 1, 1986. This has in- creased refineries' interest in the use of alternative low-cost octane boosters, par- ticularly light alkanes and methanol and/or ethanol alcohols blended with gasoline. Figure 11 shows the recent upward trend in the RVP that has resulted from these industry trends (Ford Motor Co. 1985b). Through 1980, the average RVP of the fuels sampled stayed reasonably close to the specification ofthe certification fuel (9.0 psi RVP). The increase since 1980 results from the petroleum industry's use of more light stock, such as butanes, in the gasoline. Historically, the petroleum industry has favored adding light HCs to gasoline for economic reasons. Large quantities of bu

John H. Johnson 55 2.5 2.0 E - 1.5- z o en ~ 1.0 I 0.5 O - ,~ CARBURETED VEHICLES FUEL-INJECTED VEHICLES 10 ...... Refueling 11 RVP (psi) Figure 12. Contribution to total HC emissions from various routes and processes as function of gasoline RVP. Evaporative emissions are derived from nonoperating vehicles parked overnight (diurnal); recently turned off, nonoperating vehicles (hot soak); and vehicles at filling stations receiving fuel (refueling). Exhaust emissions are derived from the tailpipe. (Adapted with permission from Stebar et al. 1985, and the Society of Automotive Engineers, Inc.) sane, a volatile HC, are produced during the refining of crude oil and natural gas. Butane has a high research octane number (about 94), and it is a good substitute for Pb in gasoline blending. The addition of bu- tane increases the "front end" volatility of a gasoline. High fuel volatility increases auto- motive evaporative emissions and increases vapor losses from fuel tanks by displacement during refueling (Stebar et al. 1985~. Stebar and coworkers (1985) analyzed a large data base (267 cars from 1978 to 1985 with 141 from 1981) to develop figure 12. The figure illustrates the importance of different HC emission routes and the con- tributions via individual routes to total vehicle HC emissions for carbureted as well as fuel-injected cars. They observed that: 1. Evaporative emissions (primarily diurnal losses; are the major contributor to the increase in vehicle HC emissions with increase in R VP. 2. Hot soak emissions (particularly with car- bureted carsJ are a larger contributor to HC emissions than are diurnal losses. 3. Reveling and exhaust HC emissions have low sensitivity to changes in RVP. 4. Exhaust emissions are the largest con  tributor to total HC emissions and constitute about the same proportion of the total for both carbureted and fi~el-injected cars. At 12 R VP, exhaust emissions represent about half of total HC emissions for both types of engines. Furey (1985) measured the vapor pres- sures and distillation characteristics of a large number of gasoline/alcohol and gaso- line/ether fuel blends. In that study, the max- imum increase in RVP above that of gasoline ranged from 0.2 psi for tert-butyl alcohol to 3.4 psi for methanol. As little as 0.25 percent methanol, ethanol, and OxinolTM 50 (a 1:1 mixture of methanol and gasoline-grade tert- butyl alcohol) was found to produce measur- able increases in RVP. The EPA estimates that the difference in volatility between the certification fuel and commercial gasoline is responsible for about half of the evaporative emissions from late-model light-duty vehicles and that this trend will continue, as shown in figure 13, if no action is taken to change it (Schwarz 1985~. The Coordinating Research Council-Air Pollution Research Advisory Committee (CRC-APRAC) is also investigating an

56 Automotive Emissions en 75 Cal I O 50 IIJ - Evaporative canister disconnects ~' ' '''"'"""'"""""'"""""""'"""""""""'"''''''''''''''''' ............... o 1980 1985 1990 YEAR 1 995 2000 Maintenance, ~ other tampering Well-maintained cars not meeting evaporative standards Figure 13. Predicted trend of the fraction of vehicles meeting evaporative emission standards and the reasons why the remaining fraction does not meet the standards. Because the RVP of commercial gasoline is different from the RVP of certification fuel, a significant percentage of vehicles will not meet the standard for this reason alone. (Adapted with permission from Schwarz 1985, and the Air Pollution Control Association.) other important gasoline fuel issue ben- zene emissions. Their preliminary findings from testing specially blended fuels in five late-model cars with three-way catalysts show that the benzene fraction of exhaust HCs increases with increasing benzene con- tent and aromaticity. In refueling and evap- orative emissions the benzene fraction in- creases with benzene content but not with aromaticity (Coordinating Research Coun- cil-Air Pollution Research Advisory Com- mittee 1985~. Fuel Usage Trends Although total gasoline usage has moved slowly upward during the past three years (from 6.5 million barrels per day (MMB/D) to 6.8 MMB/D by 1985), this trend may be temporary. The U. S. Depart- ment of Energy (1985) projects that gaso- line demand will turn downward in the balance of the 1980s and remain flat in the early 1990s as shown in figure 14. By 199.5, total gasoline consumption is projected to be 6.1 MMB/D (8.1 percent below 1983 levels). This number could be somewhat higher if the U.S. Department of Trans portation establishes the post-1988 fuel economy standards at 26 mpg. Total diesel highway fuel demand will continue to grow over the next two dec- ades primarily because of increased use of diesel engines in heavy-duty vehicles. Total highway diesel fuel usage is projected to rise 30 percent from 1.0 MMB/D in 1982 to 1.3 MMB/D in 1995, as shown in figure 14. Figure 15 shows the breakdown of  projected fuel usage by application includ- ing off-highway usage (U.S. Department of Energy 1985~. Methanol-Fueled Vehicles From an energy perspective, methanol is one of the most promising long-term alter- native fuels for motor vehicles. It can be made from natural gas now and from coal later. One major practical problem is that motor vehicle consumption for a fuel has to reach 10 percent of the present market to create an economically viable free market (Society of Automotive Engineers/U. S. Department of Energy 1985~. For the use of methanol to become widespread, it should be competitive in price with gasoline. Gas

John H. Johnson 9.0 ~ 8.0 ~ 6.0- G 5.0- LL G ~ 4.0 O- 1985 1990 1995 2000 YEAR Figure 14. Projected motor fuel consumption by fuel type. (Adapted from the U.S. Department of Energy 1985.) oline prices would probably have to exceed $1.50/gal (1985 dollars) for a significant period of time to provide the necessary confidence for investors in methanol proc- essing facilities and car buyers (Sobey 1985~. The EPA is, overall, encouraging the use of methanol as outlined by Gray (1985~. Spark-Ignition Engines for Passenger Cars. The technology for methanol-fueled vehicles exists and demonstration fleets have been tested. A summary of the emission results obtained to date for the 1983 Ford Escort fleet is given by Nichols and Norbeck (1985~. Overall the vehicles averaged 6,800 miles with a range of 3,100 to 20,300 miles. The average formaldehyde emission rate var- ied from 54 mg/mi to 79 mg/mi and ac- counted for 7.0 to 8.8 percent of the reactive HC mass on a mole-of-carbon basis. The formaldehyde as a percent of reactive HC in the exhaust for any individual vehicle ranged between 5.0 and 17.9 percent. A recent EPA summary of methanol emissions data has been documented by Alson (1985~. The HC emissions are largely 57 9.0 8.0 60 cr IIJ In 5.0 IL ~ 4.0 z o  ' 3.0 - 2.0 1 n o 1985 1990 1995 2000 YEAR Figure 15. Projected motor fuel consumption by vehicle type. (Adapted from the U.S. Department of Energy 1985.) methanol, and the aldehydes are nearly all formaldehyde. Figure 16 is a summary of formaldehyde emission data (using the FTP) comparing methanol-, diesel-, and gasoline-powered vehicles, the latter with three-way catalysts, oxidation catalysts, and no catalyst (Arson 1985~. Methanol- powered vehicles have higher formalde- hyde emissions than diesel-powered or cat- alyst-controlled gasoline-powered vehicles. 50 10 O O In ._ ~ cats O ~ t 1 METHANOL DIESEL GASOLINE Figure 16. Comparison of formaldehyde emissions from methanol-, diesel-, and gasoline-powered vehi- cles, the latter with three-way catalysts, oxidation catalysts, and no catalyst. (Adapted from Alson 1985.)

58 Automotive Emissions EGO sensor Fuel sensor ~ ~ ~ ·~ ~ = ~ ~ 1.6 L EFI engine ~ EEC IV ~ Methanol resistive fuel system Figure 17. Schematic of Ford Escort modified to use a flexible fuel system installation. Electronic en- gine controller (EEC); electronic fuel injection (EFI); exhaust gas oxygen (EGO). (Adapted with permis- sion from Wineland 1985, and the Ford Motor Co.) Ford has recently discussed the concept of a methanol/gasoline flexible fuel system that would accept either methanol or gas- oline. An electronic fuel-injected Escort was modified to use an optical fuel sensor tor determining the methanol/gasoline mixture ratio. The sensor output is con- tinuously processed by the electronic en- gine controller which optimizes fuel quan- tity and spark timing in response to the methanol/gasoline mixture ratio. This sys- tem was tested on a 1983 Escort-Lynx having a 1. 6-liter electronic fuel injec- tion engine in a production vehicle. The vehicle used the production engine com- pression ratio of 9.0, while the fuel tank, fuel filter, and the electric fuel pump were replaced with parts that methanol would not corrode. Figure 17 shows a schematic of the system in the vehicle (Wineland 1985~. Spark-Assisted, Compression-lgnition and Stratified-Charge, Spark-Ignition Buses. Methanol is considered to be a good choice as an alternative fuel for buses for several reasons. First, buses are usually a fleet operation so that methanol fuel distribution should be significantly easier than in the consumer market. Second, the particulate and odor emissions are less than those of the diesel engines it would replace. Third, the use of methanol should improve the reactivity of the exhaust, although the methanol and formaldehyde emissions could be a problem if control systems are not properly developed and maintained. Lipari and Keski-Hynnila (1985) studied the effect of a catalyst on formaldehyde emissions of a methanol-fueled, two-stroke diesel bus engine and found that even with this catalyst, emissions were still higher than those from conventional diesels in the steady-state 13-mode cycle. Additional data are needed for the transient FTP cycle and for light-load low-temperature opera-  tion, since the production of formaldehyde across the catalyst could occur under cer- tain operating conditions. Areas in Need of Additional Research. Methods for accurately measuring HC emissions from methanol-fueled vehicles are lacking. The HC unburnt fuel of a neat (100 percent) methanol vehicle is basically methanol. Measurements of HC and form- aldehyde concentrations have not been de- veloped yet that can show how high the individual excursions are under accelera- tion, deceleration, and other transient con- ditions. There is a particular lack of data taken under light-load or idling conditions, especially of operation at low temperatures. Data show that the NOx concentration can increase as exhaust from methanol-fueled vehicles passes across a catalyst. Further work needs to be done to understand this effect and to control it properly in the field or determine if it is merely a measurement problem. There is a need to study the worst-case dispersion situations outlined by Harvey et al. (1984) using these new methanol and formaldehyde concentration data. Similar experimental field studies with real-time instrumentation should also be gathered so as to ensure that methanol technology is safe in the hands of the consumer. In these latter studies, misfuel- ing and tampering should be monitored and their effects measured, for we know they occur in gasoline-powered vehicles. Recommendation 4. Formaldehyde Measurements. Real-time measurements of formaldehyde concentration should be per- formed under transient and extreme condi- tions such as acceleration and deceleration, low temperature, light load, and extended idling with restricted ventilation. This re- search work should be done with and with- out catalysts since worst-case conditions in the field will occur with catalysts removed.

John H. Johnson 59 s2 r Cetane number 50 lo: m 48 of is 111 46 44 42 ; 40 11,1,,,1,,,1,,,1,,,1,,,1 60 64 68 72 76 80 84 - - it, \ am b YEAR ~ 620 - 111 G Ul O s40 z - o m a) 600 580 560 520 Boo 480 -,,, 1 .,. 1,,, 1,,, 1,,, 1 ,,, 1 60 64 68 72 76 80 84 90% point o oo OCP 0 / - YEAR Figure 18. Historical trends of diesel fuel properties. Curve represents data from a DOE survey of type T-T diesel fuel. Open circles represent data from a MVMA survey of No. 2 diesel fuel. (Adapted with permission from Wade end Jones 1984, and the Soci- ety of Automotive Engineers, Inc.) Similar measurements should also be made on bus engines. Trends in Diesel Fuel Properties  Recent trends in diesel fuel properties have an adverse effect on particulate emissions. They make it harder to meet stringent particulate emission standards for cars and trucks (0.2 g/mi in 1987 for cars and 0.6 g/bhp-hr in 1988, 0.25 g/bhp-hr in 1991 for trucks) because the EPA certification is based on typical in-use fuels. An automo- tive quality No. 2 diesel fuel with low sul- fur and low aromatics is necessary if low particulate emissions are to be achieved (Weaver et al. 1986~. Two important fuel characteristics affecting diesel engine emis- sions have been deteriorating in recent years: the cetane number has been falling and 90 percent boiling point has been rising, as shown in figure 18 (Wade end Jones 1984~. ~ Recommendation 5. Automotive Qual- ity No. 2 Diesel Fuel. An automotive qual- ity No. 2 diesel fuel with low sulfur and low aromatics is necessary if low particu- late emissions are to be achieved. Research should be undertaken in cooperation with the automotive and petroleum industries to decide on effective and economical cetane number, sulfur and aromatic content, and 90 percent boiling point temperature spec- ification limits for automotive quality No. 2 diesel fuel, and to formulate fuels that meet the specifications. The need for this research becomes more urgent as diesel fuel usage continues to increase. Improved emissions and more control options require a quality low-sulfur diesel fuel. Refueling Emissions The basic source of HC emissions associ- ated with the vehicle refueling process is the vapors contained in vehicle fuel tanks that are displaced by gasoline during refuel- ing operations. However, additional emis- sions are associated with vehicle refueling operations as the result of "breathing losses" from underground storage tanks at gasoline service stations. Stage I (delivery of gasoline to station) vapor recovery is approximately 95 percent efficient (Austin and Rubenstein 1985~. At Stage II (dispensing of gasoline to vehicle fuel tanks) vapor recovery, gasoline vapors are collected at the vehicle fillpipe opening using a nozzle spout. The nozzle is also equipped with a vapor passage in the body of the nozzle that connects the annu- lar space between the spout and the boot to the vapor space in the underground storage tank (Austin and Rubenstein 1985~. A Stage II system reduces fillpipe emissions by 85-95 percent. Such systems are being used successfully in California. The automotive industry, the petroleum industry, and the EPA are debating whether refueling losses should be controlled by on- board vehicle systems or by Stage II systems (Austin and Rubenstein 1985; Schwarz 1985~.

60 Automotive Emissions Table 4. General Fuel Additive Classification and Typical Bulk Treatment Ranges Use Treatment Ranges Additive Type Classification Gasoline Diesel PTBa ppm PTB ~ppm Amine detergent Performance 3-30 12-120 10-60 33-200 Polymeric dispersant Performance 5-150 20-600 10-100 33-330 Fluidizer oils Performance 50-250 200-1,000 NA Antiicers Performance 015 16-60 NA Combustion modifiersb Performance (Up to 2.5 g/gal) 400-1,000 1,300-3,300 Corrosion inhibitors Distribution 1-10 040 1-10 3-33 Antioxidant Quality 3-5 12-20 2-8 7-26 Metal deactivator Quality 1-4 016 1-4 016 Demulsifiers Distribution 0.1-2.5 0.010 0.1-2.5 0.3-S Flow improver Performance NA 15-150 50-500 a Pounds of fuel additive per 1,000 barrels of fuel where 1 barrel = 42 gal. b Combustion modifiers = octane (gasoline) or cetane (diesel) number improvers. NOTE: NA = not applicable. SOURCE: Adapted with permission from Tupa and Doren 1984, and the Society of Automotive Engineers, Inc. By use of onboard control systems, vapors displaced from the vehicle tank are vented to an enlarged canister where they are absorbed and subsequently purged into the engine (Austin and Rubenstein 1985; Schwarz 1985~. A separate canister to control refueling emis- sions or an enlarged evaporative canister could be used. In a further detailed analysis of Stage II and onboard control, Austin and Ruben- stein (1985) reached the following general conclusion: "the implementation of Stage II controls is a clearly superior alternative to the onboard control concept." Their spe- cific reasons for this conclusion were: 1. Stage II controls have been proven In California and they can achieve about 85 percent control. 2. Stage II controls are the more cost-ef- fective, that is, $0.21/lb HC are reduced ver- sus $0.66 to $2.25/lb HC depending on whether the EPA's or Ford's cost estimate is used for onboard control. The onboard sys- tems can be made more cost-e~ective with additional evaporative emissions control. 3. Stage II controls give better short-term control because of lead time, and vehicle turnover due to replacement, among others as shown by Austin and Rubenstein (1985~. Recommendation 6. Evaporative Emis- sions Control. Some combination of field RVP control along with a test fuel with typical RVP (or calculation corrections for RVP differences) should be developed. Controlling RVP in motor gasoline, an approach successfully applied in California, is needed generally for controlling field evaporative emissions. The question of whether car manufacturers should be test- ing with a worst-case RVP test fuel or a  typical fuel needs further study. Additives Additives are used to improve engine per- formance and durability and to ensure that fuel specifications and quality are main- tained during transport and storage. They are an integral part of today's fuels. Tupa and Doren (1984) discuss in great detail the specific functions and benefits of additives, typical use levels, and test methods for evaluation. Generic types of additives and their uses are shown in table 4 along with general levels of additive treatment for the various types of additives. The variety of chemical compounds used in gasolines to- day are listed in table 5 and table 6 (Tupa and Doren 1984~. How additives may affect control tech- nologies needs additional research. Knowl- edge of the size distribution of particles from diesel particulate traps and the metal species they contain. Data on operation

John H. Johnson 61 Table 5. Chemicals Typically Used for Gasoline Additives Additive Type Chemicals Amine detergent Polymeric dispersants Fluidizer oils Amines Alkanol amines Amides Amido-amines Imidazolines Alkenyl succinimides Hydrocarbyl amines Polyether amine Selected mineral oil Thermally stable polyolefin (polypropylene) of mod- erate molecular weight (800-1,000) and narrow molecular weight distri- bution Ester-type synthetic lubri- cant Combustion modifiers Tetraethyl lead (TEL) (unlit in 1P~APA {li`>lc Antiicers Corrosion inhibitors \ , , ^ ^ ~ ,. A, which are less than 35% of gasoline sold today) Tetra m eth yl lead (TM L) Methyl cy clo p en ta dien yl Manganese tricarbonyla (MMT) Alkenyl succinates and amine salts Monocarboxylic acids and amine salts Imidazolines and carboxylic  acid salts Amine alkylorthophos phates Ethoxylated alkyl phenols Alkenyl succinic acids, es ters, and amine salts Dimer acid and other car boxylic acids and amine salts Mixed alkyl orthophospho ric acids and amine salts Alkyl phosphoric acids and amine salts Aryl sulfonic acids and amine salts Mannich amines Carboxylic acid salts of Mannich amines Demulsifiers Long-chain alkylphenols Long-chain alcohols Long-chain carboxylic acids ~Long-chain amines (Table continued next column.) Table 5. Continued Additive Type Antioxidants Chemicals Metal deactivators Alkyl- or aryl-substituted phenolenediamines Alkyl- or aryl-substituted aminophenols Alkyl- or aryl-substituted phenols Fuel treated with N,N'-dis- alicylidene-1, 2-propane Diamine which produces copper chelate a These additives are not used in unleaded gasoline, and MMT was specifically banned by the 1977 Clean Air Act Amendments. SOURCE: Adapted with permission from Tupa and Doren 1984, and the Society of Automotive Engi- neers, Inc. with as well as without the working traps are needed since tampering of control de- vices can occur in the field. The effect of the additive compounds that plug the trap pores also needs further study. Recommendation 7. Diesel Fuel Ad- ditives. Data should be obtained about the size distribution of particles in diesel ex- haust and about the metal species they con- tain, with and without a particulate trap, with a diesel fuel containing a typical addi- tive under consideration for production use. Data on the HCs bound to the particles and the vapor-phase HCs should also be ob-  tained. Methods for Measuring the Unregulated Pollutants* The unregulated pollutants in automotive exhaust have been measured with varying degrees of sophistication for the past 20 years. Interest in a particular pollutant var- ies as studies of its potential health effects are reported; benzotaipyrene represents a good example. Since most of the unregu- lated pollutants are present only in small amounts in exhaust (in the parts-per *This section was written by David Leddy, Michigan Technological University.

62 million [ppm] range or less) and very small amounts in the ambient air (in the parts per-billion [ppb] range or less) their amount or concentration is measured only with great difficulty and usually at high expense. There is little doubt that there is a need to balance the degree of diff~cul ty, the cost, and the sensitivity against the real value the procedure produces in assessing health effects or engine perform ance. Sampling Since most of these pollutants are found at low concentrations, nearly all methods of analysis call for collecting a sample over an extended time interval and concentrating it before analysis. Samples are frequently col lected by the use of impingers, filters (Evans 1980; Perez et al. 1980; Gorse and Salmeen 1982; Gross et al. 1982; Fox 1985), and solid sorbents (Hampton et al. 1982; Fox 1985~. Sampling of the exhaust may be more important in determining the value of the analysts 'then the actual measurement itself. For example, there is every reason to be lieve that during the sampling of particu lates on a filter, chemical reactions take place between the organic compounds in Demulsifiers 'the particulates and gaseous or aerosol compounds such as nitric acid (HNO3), NO2, and sulfuric acid. These reactions produce the so-called "artifacts of sam pling" that are of concern to all who work in this field (Lee et al. 1980; Perez et al. 1980; Pierson et al. 1980; Gorse and Sal meen 1982; Herr et al. 1982; Risby and Lestz 1983~. One of the possible sampling artifacts of greatest concern is the forma tion of the biologically active compounds nitropyrene and nitrobenz~a~pyrene from the reaction of NO2 with the relatively innocuous compounds pyrene and ben zota~pyrene, respectively (Gibson et al. 1980; Schuetzle et al. 1980~. Other artifacts of concern are the formation of HNO3 and sulfuric acid on the surface of the sampling material. The effects of artifact formation can be minimized by reducing the length of time a filter is exposed to the exhaust stream to the minimum required to collect Deposit modifiers Automotive Emissions Table 6. Chemicals Typically Used as Diesel Additives Additive Type Chemicals Detergents Polymeric dispersants Combustion modifiers Cetane improvers Same as for gasoline (table 5) Same as for gasoline  (table 5) Alkyl nitrates and nitrites Nitro and nitroso compounds Peroxides Combustion catalysts Organo compounds of barium, calcium, manganese, and iron for changing output particulate emissions Organo compounds of manganese, copper, lead, cerium, or com binations of above metals in particulate traps to reduce regen eration temperatures Barium, calcium, or manganese Ethylene vinyl acetate polymers Chlorinated hydrocarbons Polyolefins Same as for gasoline (table 5) Same as for gasoline (table 5) Most fuels in United States Borate esters Quarternary ammonium salts of salicylic acids Diamine complexes of nickel Organo barium compounds Glycol ethers Same as for gasoline (table 5) Tertiary amines Imidazolines Tertiary alkyl primary amines Same as for gasoline (table 5) Corrosion inhibitors Biocides (to inhibit bac- teria growth at water/ fuel interface) Stabilizers An tio xi dan ts Metal deactivators SOURCE: Adapted with permission from Tupa and Doren 1984, and the Society of Automotive Engi- neers, Inc.

John H..Johnson 63 Table 7. Summary of Analytical Methods for Characterizing Unregulated Emissions from Spark- Ignition and Compression-Ignition Engines E. . mission Sampling Impinger/DNPH Analytical Method Reference Aldehydes HCN, cyanogen Organic sulfides Ammonia Organic amines NOX SO2 HCs Phenols N-nitrosodi- methylamine Benzotaipyrene Sulfates Nitric acid Metals GC, HPLC, calorimetry Impinger/KOH Tenax traps Impinger/H2SO4 Impinger/H2SO4 Tedlar bag Impinger/H202 Real-time Tedlar bag, trap Impinger/KOH Impinger/NaOH Filter, trap, poly- urethane foam Filter Filter Filter GC/ECD GC/FID IC GC/NPD GC/ECD Colorimetry, IC Electrochemical, fluores- cence spectroscopy; 2nd derivative spectroscopy GC/FID GC/FID GC/MS GC/MS, HPLC  Colorimetry, IC IC Atomic absorption, x-ray fluorescence, emission spectroscopy Perez et al. (1980, 1984) Fox (1985) Cadle et al. (1979); Perez et al. (1980) Cadle et al. (1979); Perez et al. (1980) Cadle et al. (1979); Perez et al. (1980) Cadle et al. (1979); Mulawa and Cadle (1979) Cadle et al. (1979) Perez et al. (1980) Fox (1985) Perez et al. (1980); Fox (1985) Perez et al. (1980) Krost et al. (1982) Schuetzle et al. (1980); Eisenberg (1983); Eisenberg et al. (1984) Perez et al. (1980); Schuetzle et al. (1980) Okamoto et al. (1983) Perez et al. (1980) NOTE: DNPH = 2,tdinitrophenylhydrazine; GC = gas chromatography; GC/ECD = GC with electron- capture detection; GC/FID = GC with flame ionization detection; GC/NPD = GC with nitrogen-phosphorus detection; GC/MS = GC/mass spectrometry; HCN = hydrogen cyanide; H2O2 = hydrogen peroxide; HPLC = high performance liquid chromatography; H2SO4 = qu]filric arid TV. = ion rbrnm~tr~s:rr~nh`7 K~M = nr~t~ccil~m hydroxide; and NaOH = sodium hydroxide. a suitable sample, by using inert materials for filter construction, and by cooling and diluting the exhaust stream prior to sample collection. Analytical Methods Not every analytical method used for char . . . . . . . acter1z1ng emissions trom spar. `-1gn1t1on engines is applicable for analysis of emis . . . . . . sloes trom compress1on-lgn1t1on engines, because of interference from combustion products found in compression-ignition engines. Diesel engines produce higher lev- els of particulates, NOx, sulfur oxides (SON and certain HCs, all of which can interfere with one or more of the analyses that are commonly used on spark-ignition engine emissions. Some real-time moni- toring techniques based on the absorp- tion of light fail when applied to diesel exhaust analysis, either because of scatter -- =- -r -- ~ ~ ~r ~ ing of light by suspended particulates or absorption of light by aromatic HCs present in the gaseous phase. Electrochem- ical methods are affected because particu- lates foul the membranes and electrode surfaces used in the measuring cells. Apply- ing some of the methods used for con- tinuous monitoring of chemical species in ambient air is even more difficult when one considers that spark-ignition as well  . . . . . as compress1on-1gn1t1on engines generate interfering species that affect the sensi- tivity, accuracy, and repeatability of the analyses. The analytical methods used to measure concentration or amount of unregulated pollutants are summarized in table 7. It should be emphasized that these are the analytical methods presently used in labo- ratories where measurements are being made on a regular basis for judging engine performance. They meet current require

64 Automotive Emissions meets but will not necessarily meet the requirements of the future. We need to find out which of the unreg- ulated pollutants must be measured to eval- uate advanced technology for the control of emissions from gasoline as well as diesel engines. It would be folly to measure the concentrations of pollutants just because they are there. Present methods of analysis are so tedious, expensive, and unreproduc- ible that unnecessary analyses are to be avoided whenever possible. Areas in Need of Additional Research. Gaps in unregulated emission measurement methods center on the lack of real-time measurement methods that have the sensi- tivity required for producing results at moderate costs. If advanced emission con- trol technology is to be studied with tran- sient-cycle test protocols, these real-time techniques are necessary. Real-time measurements based upon p~- ezoelectric devices, tunable diode laser sys- tems, thermal lens spectroscopy, long-path differential optical absorption spectros- copy, ultraviolet fluorescence spectros- copy, and differential absorption lidar have been reported by Fox (1985~. The studies reported in these cases are normally of ambient air with no concern for interfer- ences that may be present in gasoline and diesel exhaust. Pitts et al. (1984) reported the measurement of gaseous HNO3, NOD, formaldehyde, SO2, and benzaldehyde in the exhaust of light-duty vehicles, using an instrument that coupled a multiple reflec- tion cell to a differential optical absorption spectrometer. The techniques hold promise for the future and should be explored in more extensive studies. Recommendation 8. Emissions Mea- surement Methods. Studies should begin immediately with an evaluation of the best available emissions data on engines operat- ing with and without emission control de- vices, to determine which of the unregu- lated pollutants really pose a potential threat to human health. Other unregulated pollutants might be added to this list if their concentrations reflect engine or emission control device performance. Next, every effort should be made to improve the ana- lytical procedures presently used to mea- sure the concentrations of those pollutants, to the point where they can be readily carried out by technicians. This may re- quire that packaged sets of reagents and equipment be marketed for a specific anal- ysis. For example, prepacked traps might be available for collecting gaseous HC prior to thermal desorption onto a gas chroma-  tograph with a specified capillary column for the analysis of specific HCs at predeter- mined conditions. Current Regulated and Unregulatec! Emissions The main focus of this section is unregu- lated organic emissions, for significant data on regulated emissions and non-organic unregulated emissions from in-use vehicles have been presented already. MOBILES, a computer model discussed earlier, is the best source of data about regulated emis- sions since the EPA analyzes all manufac- turers' data and develops sales-weighted emission factors (U. S. Environmental Pro- tection Agency 1985~. Emission factors for regulated pollutants, based on California Air Resources Board (1980) data, are also available. Regulated Emissions Table 8 from the National Research Coun- cil (1983b) shows a summary of the regu- lated emissions from light-duty vehicles. Imposing the HC, CO, and NO',. standards has resulted in 84, 79, and 56 percent reductions, respectively, in 50, 000-mi emissions from gasoline-powered, spark- ignition vehicles. Unregulated Emissions Gas-Phase Hydrocarbons. The compo- nents detected as gas-phase HCs are listed in table 9 (National Research Council 1983b). In another report, the National Research Council (1983a, appendix A) prepared an extensive list of vapor-phase compounds

John H. Johnson 65 Table 8. Exhaust Emission Ratesa for Light- Duty Gasoline-Powered Vehicles Emission Model Component Year Zero-Mile 50,000-Mile E· ' TO ' ' mlsslon emlsslon Rate Rateb (g/mi) (g/mi) HCs NOB Pre-1968 1968-1969 1970-1971 1972-1974 1975-1979 1980 1981 1982+ Pre-1968 1968-1969 1970-1971 1972-1974 1975-1979 1980 1981 1982 1983+ Pre-1968 1968-1972 1973-1974 1975-1976 1977-1979 1980 1981 + 7.25 4.43 3.00 3.36 1.29 0.29 0.39 0.39 78.27 56.34  42.17 40.78 20.16 6.14 5.60 5.21 5.00 3.44 4.35 2.87 2.43 1.69 1.56 0.75 8.15 5.68 4.85 4.21 2.74 1.74 1.34 1.34 89.52 69.09 57.82 52.98 34.46 20.44 19.35 19.01 18.80 3.44 4.35 3.07 2.63 2.19 2.06 1.50 a Emission rates are for low-altitude 49-state vehicles. High-altitude and California emission rates are differ- ent. b The 50,000-mile emission rates are calculated from zero-mile rate by addition of term that takes account of EPA-projected deterioration rate of vehicle com- bustion and emission-control systems. SOURCE: Adapted with permission from the Na- tional Research Council 1983b. in both diesel- and gasoline-powered vehi- cles by reviewing 250 papers in the litera- ture. Diesel Exhaust Particulate. Diesel ex- haust particulate material has been the sub- ject of extensive study in the past five years. It is typically about 25 percent extractable into organic solvents, although different vehicles may have extractable fractions of ~90 percent, depending to some extent on operating conditions. More than half the extractable material is aliphatic HC of 1~35 carbon atoms, alkyl-substituted ben Table 9. Unregulated Gaseous Hydrocarbons Emitted from Vehicles  All n-alkanes from e-butane through n-hexacosane Four methyl-substituted butanes Ten methyl- and ethyl-substituted pentanes and 11 cyclopentanes Eleven methyl- and ethyl-substituted hexanes and 35 cyclohexanes Fifteen methyl- and ethyl-substituted heptanes Five methyl-substituted octanes One methyl-substituted nonane One methyl-substituted decane One methyl-substituted undecane Decalin and two methyl-substituted decalins Two Can alkanes Eleven Cal alkanes Nine Cal alkanes Thirteen C~3 alkanes Eleven C~4 alkanes Eight C's alkanes Eight C'6 alkanes Five C~7 alkanes Three C,~ alkanes Seven methyl-substituted butenes and two methyl butadienes Eighteen pentenes and pentadiene Fourteen hexenes Six heptenes Four octenes Decene and dodecene through heneicosene Seven cyclic olefins Seventy-one alkyl-substituted benzenes Eight styrenes and the three xylenes Fourteen indans and three indenes Twenty-eight alkyl-substituted naphthalenes Three alkylthiophenes and two benzothiophenes Two alkylsulfides and one alkylamine Six nonaromatic alcohols and eight aromatic alcohols Eighteen aliphatic and aromatic aldehydes Six furans, 17 ketones, and six organic acids SOURCE: Adapted with permission from the Na- tional Research Council 1983b. zones, and derivatives of the polycyclic aromatic hydrocarbons (PAM) such as ke- tones, carboxaldehydes, acid anhydrides, hydroxy compounds, quinones, nitrates, and carboxylic acids. There are also hetero- cyclic compounds containing sulfur, nitro- gen, and oxygen atoms within the aromatic ring. The alkyl-substituted PAHs and PAH derivatives tend to be more abundant than the parent PAH compound (National Re- search Council 1983b).

66 Automotive Emissions Table 10. Qualitative Analysis of Nonpolar and Moderately Polar Fractions of Diesel Particulate Extract Compound Approximate Concentration in Oldsmobile Extract (ppm) Nonpolar fractions Phenanthrenes and anthracene Methylphenanthrenes and methylanthracenes Dimethylphenanthrenes and dimethylanthracenes Pyrene Fluoranthene Methylpyrenes and methylfluoranthenes Chrysene Cyclopenta|cdlpyrene Benzotghi]fluoranthene Benz~ajanthracene Benzota~pyrene Other PAHs, heterocyclics HCs and alkylbenzenes Total nonpolar fractions Moderately polar fractions PAH ketones Fluorenones Methylfluorenones Dimethylfluorenones Anthrones and phenanthrones Methylanthrones and methylphenanthrones D imeth ylan thrones and dim eth ylp hen anthro nes Fluoranthones and pyrones Benzanthrones Xanthones Methylxanthones Thioxanthones Methylthioxanthones Total PAH carboxaldehydes Fluorene carboxaldehydes Methyl fluorene carboxaldehydes Phenanthrene and anthracene carboxaldehydes Methylanthracene and methylphenanthrene carboxaldehydes Dimethylanthracene and dimethylphenanthrene carboxaldehydes Benz~ajanthracene, chrysene, and triphenylene carboxaldehydes Naphthalene dicarboxaldehydes Dimethylnaphthalene carboxaldehydes Trimethyluaphthalene carboxaldehydes Pyrene and fluoranthene carboxaldehydes Xanthene carboxaldehydes Dibenzofuran carboxaldehydes Total PAH acid anhydrides Naphthalene dicarboxylic acid anhydrides Methylnaphthalene dicarboxylic acid anhydrides Dimethylnaphthalene dicarboxylic acid anhydrides Anthracene and phenanthrene dicarboxylic acid anhydrides Total (Table continued next page.) 600 1,400 3,000  1,700 1,400 800 100 20 100 500 40 30,000 500,000 539,700 4,000 400 200 1,600 1,600 1,300 1,200 200 300 200 1,600 900 13,500 1,600 400 2,600 1,600 400 400 300 300 1,000 1,600 600 400 1 1,200 3,000 1,000 500 600 5,100

John H. Johnson 67 Table 10. Continued Compound Approximate Concentration in Oldsmobile Extract (ppm) Hydroxy PAHs Hydroxyfluorene Methylhydroxyfluorene Dimethylhydroxyfluorene Hydroxyanthracenes and hydroxyphenanthrenes Hydroxymethylanthracenes and hydroxymethylphenanthrenes Hydroxydimethylanthracenes and hydroxydimethylphenanthrenes Hydroxyfluorenone Hydroxyxanthone Hydroxyxanthene Total PAH quinones Fluorene quinones Methylfluorene quinones Dimethylfluorene quinones Anthracene and phenanthrene quinones Methylanthracene and methylphenanthrene quinones Fluoranthene and pyrene quinones Naphthot1,8-cdlpyrene 1,3-dione Total Nitro-PAHs Nitrofluorenes Nitroanthracenes and nitrophenanthrenes Nitrofluoranthenes Nitropyrenes Methylnitropyrenes and methylnitrofluoranthenes Total Other oxygenated PAHs PAH carryover from nonpolar fraction Phthalates, HC contaminants Total, moderately polar fractions 1,400 400 1,500 600 900 1,300 2,000 1,300 1,000 10,400 700 600 500 1,900 2,000 200 600 6,500 30 70 10 150 20 280  8,000 6,000 30,000 91,000 SOURCE: Adapted with permission from the National Research Council 1983b. The particulate-extract in the high-per- formance liquid chromatograph (HPLC) eluent can be separated into nonpolar, moderately polar, and highly polar frac- tions. The fractions can then be further analyzed by gas chromatography/mass spectrometry (GC/MS). Table 10 lists the results of such an analysis of the nonpolar and moderately polar fractions of a partic- ulate extract from an Oldsmobile diesel vehicle, including the approximate extract concentrations for this particular vehicle. The highly polar fraction has not been fully characterized. It contains the PAH carbox- ylic acids, acid anhydrides, and probably sulfonates and other highly polar species (National Research Council 1983b). Most (75 percent) of the direct bacterial mutagenicity resides in the moderately po- lar fraction. The remaining direct mutage- nicity is in the highly polar fraction. These aspects are discussed further in the National Research Council's report (1983b). Over 50 chromatographic peaks of nitro- PAH compounds have been identified in diesel particulate extracts, as listed in table 11. The most abundant of the nitro-PAHs is 1-nitropyrene, ranging from 25 to 2,000 ppm in the vehicle extracts studies. The other nitro-PAHs are present at concentra- tions from below the ppm range to a few ppm. The nitropyrenes have been studied in greater detail than other PAH com- pounds. They are released in diesel and

68 Table 11. Nitroarenes in Diesel Exhaust Particulate Extracts Mononitroarenes N i tro in den e Nitroacenaphthylene Nitroacenaphthene Nitrobiphenyl Nitrofluorene Nitromethylacenaphthylene Nitromethylacenaphthene Nitromethylbiphenyl Nitroanthracene Nitrophenanthrene Nitromethylfluorene Nitromethylanthracene Nitromethylphenanthrene Nitrotrimethylnaphthalene Nitrofluoranthene Nitropyrene Nitro (C2-alkyl) anthracene Nitro (C2-alkyl) phenanthrene Nitrobenzofluorene Nitromethylfluoranthene Nitromethylpyrene Nitro (C3-alkyl) anthracene Nitro (C3-alkyl) phenanthrene Nitrochrysene Nitrobenzoanthracene Nitronaphthacene Nitrotriphenylene Nitromethylnaphthacene or nitromethylchrysene Nitromethylbenzanthracene Nitromethyltriphenylene Nitrobenzopyrene Nitroperylene Nitrobenzofluoranthene Automotive Emissions Polynitroarenes Dinitromethylnaphthalene Dinitrofluorene Dinitromethylbiphenyl Dinitrophenanthrene Dinitropyrene Trinitropyrene Trinitro (C5-alkyl) fluorene Dinitro (C6-alkyl) fluorene Dinitro(C4-alkyl)pyrene Nitro-oxyarenes Nitronaphthaquinone Nitrodihydroxynaphthalene Nitronaphthalic acid Nitrofluorenone Nitroanthrone Nitrophenanthrone Nitroanthraquinone Nitrohydroxymethylfluorene Nitrofluoranthone Nitropyrone N itro fl u o ran theneq uin one Nitropyrenequinone Nitrodimethylanthracene carboxaldehyde  N it rodi m eth yl p henan t hrene ca rb o x aldeh y de Other nitrogen compounds Benzocinnoline Methylbenzocinnoline Phenylnaphthylamine (C2-alkyl) phenyloaphthylamine SOURCE: Adapted with permission from the National Research Council 1983b. gasoline exhaust (according to particulate extracts) at rates of approximately 8.0 (diesel fuel), 0.30 (leaded gasoline), and 0.20 mg/mi (unleaded gasoline) (National Research Council 1983b). 1-Nitropyrene has been the only ni- tro-PAH detected in spark-ignition par- ticulate extracts. Very low 1-nitropyrene particulate extract concentrations have been found recently in on-road heavy- duty diesel and light-duty spark-ignition vehicles (National Research Council 1 983b). Gasoline-Powered Vehicle Refueling Hydro- carbons. Williams (1985) has reported the concentration of HCs in the breathing zone of individuals during vehicle refueling. Gas chromatographic data for gasoline and the refueling vapor indicate that only the lower molecular weight, more volatile com- pounds are emitted. Williams concluded that: 1. Vapor composition does not equal 1- . . gaso~ne compos~t~on; 2. Range of total HC concentrations varied widely with the environmental con- ditions, resulting in exposures from 8 to 3,000 ppmC; and 3. Propane, butane, and pentane provide more than 80 percent of total exposure.

John H..Iohnson 69 Areas in Need of Additional Research. To do efficient particulate control develop- ment work and to better understand emis- sion characteristics, there is a need for a fast-response real-time particulate mass measurement instrument. The tapered ele- ment oscillating microbalance (TEOM) holds the most promise, but there is a gap between what is known about its principle of operation and the reality of making its use practical for measuring par- ticulates. The other instrument gap involves ~ea- suring methanol accurately. Formaldehyde has also been identified as a potentially important unregulated pollutant that needs careful real-time measurement and control because it is generally considered to be a carcinogen. Measurement of particulate emissions from heavy-duty diesel engines using the EPA test procedures with dilution tunnels is inadequate. The current reneatabilitv of measurements is poor. Barsic (1984) showed trom a round-robin test that the root mean square of the 2-~ standard devi- ations were 76 percent of a 0.25 g/bhp-hr standard for six heavy-duty diesel engines tested in seven laboratories. For measure- ments intended to implement the 0.25 g/bhp-hr or the 0.1 g/bhp-hr standard, this variation is unacceptable. It is uncertain whether particulate emis- sion standards should be based on amount of total particulate matter, on which cur- rent standards are based, or amount of soluble organic component extracted from the particulates. The soluble organic com- ponent is the portion of the particulate that has been shown to be mutagenic and pos- sibly carcinogenic (Claxton 1983), suggest- ing that future health-related regulations should be based on this fraction. Basing standards on the soluble organic compo- nent poses the problem of separating and quantifying the specific toxic components by one of the present methods solvent extraction, vacuum sublimation, or ther- mogravimetric analysis. Variability associ- ated with the separation methods and sam- pling condition affects the mass of the soluble fraction collected, compounding the previously stated measurement vari ability problem for the total particulate matter. Present measurement methods for the collection of vapor-phase HCs from diesel engines do not collect all the compounds. Characterization of potentially toxic HCs is not possible if they cannot all be col- lected. There is need to continue the develop-  ment and use of advanced HPLC and GC/MS techniques In conjunction warn separation methods to more accurately measure the amounts of key biologically active HCs in the particulate as well as the vapor phases. The nitroaromatics are im- portant compounds whose concentrations in diesel exhaust with and without partic- ulate traps should be measured more accu- rately. There is a need to investigate and de- velop measurement methods that quantify diesel odor. Pioneering work was carried out in the late 1960s and early 1970s, but was dropped around 1978 because of the potential health effects of diesel particulate emissions. Diesel odor, along with partic- ulates, is still the typical person's percep- tion of the diesel pollutants that are of concern. There is a need to apply odor measurement methods to new engines used in light-duty and heavy-duty vehicles and advanced engines that use particulate traps or incorporate advanced high-temperature materials. Refined organic compound measure- ment is particularly important to advance the development of low-heat-rejection (or commonly called adiabatic, as an ideal goal) diesel engines because their combustion chamber wall and gas temperatures will be higher. This elevated temperature will in- crease the amount of lubricating oil appear- ing as particulate emissions and has the potential of producing reactions between the HCs and oxygen/HNO3/NOx and other such gaseous mixtures to form toxic and biologically active species. A particular need in unregulated pollu- tant characterization data for gasoline en- gines is additional nitrous acid (HNO2) data as an extension to Pitts et al. (1984~. That paper showed higher levels of HNO2 from older (1974 and earlier) light-duty

70 Automotive Emissions vehicles than from 1982 and newer cars that use three-way catalyst systems. The data show that even though the number of older cars is small, their HNO2 emission levels are so high that they may be the major source of all gaseous HNO2 from automo- tive emissions. HNO2 is a key precursor to photochemical air pollution and is also an inhalable nitrite. There is little known about how and when the nitro-PAHs are formed in the exhaust system (or the dilution tunnel) of diesel engines. Flow reactor studies with the basic species NO, NO2, CO, CON, N2, 02, S02, HCs present in the ex- haust, along with detailed engine studies that include the effects of the particles in the reactions, could help resolve this issue. Recommendation 9. PAH Measure- ments. There is a need for a program of comparative measurements of PAHs in partial-exhaust sampling systems and in full-flow dilution tunnel systems, with measurements made in the atmosphere downwind from the plume, for the pur- pose of determining how well laboratory data reflect the true composition of emis- sions into the atmosphere. Recommendation 10. Kinetics of Ni- tro-PAH Formation. Research is recom- mended to discover how and when nitro- PAHs are formed in the diesel engine exhaust system and dilution tunnel. This work can best be done by flow reactor studies of the basic gases in conjunction with detailed engine studies that include the actual HCs and particles. Recommendation 11. Particulate Mea- surement Variability. Research is required to reduce the variability in heavy-duty die- sel particulate measurements. Work needs to be undertaken to determine how to better control the parameters that influence this variability. a Recommendation 12. HC Character- ization. There is a need for research on the complete characterization of particulate- phase and gas-phase HCs in diesel exhaust. a Recommendation 13. Diesel Odor. There is a need to investigate and develop analytical methods that quantify diesel odor. This research should take advantage of the knowledge gained in the past eight years about measuring particulate-bound and vapor-phase HCs. Recommendation 14. Nitrous Acid.  Additional data should be obtained about HNO2 emissions from older gasoline- powered vehicles. The literature shows high levels of HNO2 from older cars that may be contributing significantly to in- creased photochemical smog and direct ef- fects. . . Summary of Research Recommendations HIGH PRIORITY Based upon current information, the following research studies are most likely to yield useful data. Recommendation 1 Tampering and misfueling statistics are fairly well developed but Tampering and their effect on emissions is not as well known. Therefore, work Misfiteling should be done to better characterize the effect of tampering and misfueling on emissions from vehicles and to better assess their effect on ambient pollutant concentrations.

John H..Iohnson 71 Recommendation2 There is a need for continued research on particulate control Diesel Particulate technology, including the regeneration systems, to reduce the cost Emissions and and complexity of these systems and the associated fuel economy Control penalties. Work needs to continue with various additives, substrate materials, regeneration systems, and controls to develop optimum systems that are able to decrease the diesel particulate emissions to the levels of 0.1 ~/bho-hr for heavy-duty diesels and 0.08 Mimi - ~ ---r ~~~ ~~ ' for light-duty vehicles required in California. In conjunction with this research there is a need to measure the metal species and the size distribution of the particles coming from diesel particulate traps. Recommendation 4 Real-time measurements of formaldehyde concentration should Formaldehyde be performed under transient and extreme conditions such as Measurements acceleration and deceleration, low temperature, light load, and extended idling with restricted ventilation. This research work should be done with and without catalysts since worst-case condi tions in the field will occur with catalysts removed. Similar measurements should be made on bus engines. Recommendation 7 Data should be obtained about the size distribution of particles in Diesel Fuel diesel exhaust and about the metal species they contain with and Additives without a particulate trap, with a diesel fuel containing a typical additive under consideration for production use. Data on the HCs bound to the particles and the vapor-phase HCs should also be obtained. Recommendation 10 Research is recommended to discover how and when nitro Kinetics of Nitro- PAHs are formed in the diesel engine exhaust system and dilution PAH Formation tunnel. This work can best be done by flow reactor studies of the basic gases in conjunction with detailed engine studies that include the actual HCs and particulates. Recommendation 11 Particulate Measurement Variability Research is required to reduce the variability in heavy-duty diesel particulate measurements. Work needs to be undertaken to deter- mine how to better control the parameters that influence this variability. Recommendation 12 There is a need for research on the complete characterization of HC Characterization particulate-phase and gas-phase HCs in diesel exhaust. MEDIUM PRIORITY Recommendations An improved vehicle evaporative emissions model should be Evaporative developed that is valid over various types of operating conditions Emission Model for a variety of ambient temperatures. At the same time, changes should be made in the EPA test procedure to obtain the data necessary to properly design and size the evaporative system for the high-temperature soak situation, and data should be sought that can be used in EPA's MOBILES computer model for other use patterns of cars.

72 Automotive Emissions Recommendation 5 An automotive quality No. 2 diesel fuel with low sulfur and low Automotive Quality aromatics is necessary if low particulate emissions are to be No. 2 Diesel Fuel achieved. Research should be undertaken in cooperation with the automotive and petroleum industries to decide on effective and economical cetane number, sulfur and aromatic content, and 90 percent boiling point temperature specification limits for automo tive quality No. 2 diesel fuel, and to formulate fuels that meet the specifications. The need for this research becomes more urgent as diesel fuel usage continues to increase. Improved emissions and more control options require low-sulfur diesel fuel. Recommendation 6 Some combination of field RVP control along with a test fuel Evaporative with typical RVP (or calculation corrections for RVP differences) Emissions Control should be developed. Controlling RVP in motor gasoline, an approach successfully applied in California, is needed generally for controlling field evaporative.emissions. The question of whether car manufacturers should be testing with a worst-case RVP test fuel or a typical fuel needs further study. Recommendations Studies should begin immediately with an evaluation of the Emissions best available emissions data on engines operating with and with Measurement out emission control devices, to determine which of the unregu Methods lated pollutants really pose a potential threat to human health. Other unregulated pollutants might be added to this list if their concentrations reflect engine or emission control device perform ance. Next, every effort should be made to improve the analytical procedures presently used to measure the concentrations of those pollutants, to the point where they can be readily carried out by technicians. This may require that packaged sets of reagents and equipment be marketed for a specific analysis. For example, pre packed traps might be available for collecting gaseous HCs prior to thermal desorption onto a gas chromatograph with a specified capil lary column for the analysis of specific HCs at predetermined condi tions. Recommendation 9 There is a need for a program of comparative measurements of PAH measurements PAHs in partial-exhaust sampling systems and in full-flow dilution tunnel systems, with measurements made in the atmosphere down wind from the plume, for the purpose of determining how well laboratory data reflect the true composition of emissions into the atmosphere. Recommendation 13 There is a need to investigate and develop analytical methods that Diesel Odor quantify diesel odor. This research should take advantage of the knowledge gained in the past eight years about measuring partic ulate-bound and vapor-phase HCs. Recommendation 14 Additional data should be obtained about HNO2 emissions from Nitrous Acid older gasoline-powered vehicles. The literature shows high levels of HNO2 from older cars that may be contributing significantly to increased photochemical smog and direct effects.

lohn H..Iohnson Acknowledgment I would like to thank Peter V. Woon for all of his assistance in the preparation of this paper. References Abtoff, J., Schuster H. C., and Langer, J. 1985. The trap oxidizers an emission control technology for diesel engines, Society of Automotive Engineers Paper 850015, Warrendale, Pa. Alson, J. 1985. EPA methanol vehicle emissions test programs, talk at EPA Region VI Methanol Work- shop, May 1985, Dallas, Tex. Amann, C. A. 1985. The powertrain, fuel economy and the environment, General Motors Corp. Research Publication 4949, Warren, Mich. Austin, T. C., and Rubenstein, G. S. 1985. A com- parison of refueling emissions control with onboard and Stage II systems, Society of Automotive Engi- neers Paper 851204, Warrendale, Pa. Barnes, G. J., and Donohue, R. J. 1985. A manufac- turer's view of world emissions regulations and the need for harmonization of procedures, Society of Automotive Engineers Paper 850391, Warrendale, Pa. Barsic, N. J. 1984. Variability of heavy-duty diesel engine emissions for transient and 13-mode steady- state test methods, Society of Automotive Engi- neers Paper 840346, Warrendale, Pa. Cadle, S. H., Nebel, G. J., and Williams, R. L. 1979. Measurements of unregulated emissions from Gen- eral Motors' light-duty vehicles, Society of Auto- motive Engineers Paper 70694, Warrendale, Pa. California Air Resources Board. 1980. Procedure for estimating on-road motor vehicle emissions, Sacra- mento, Calif. Claxton, L. D. 1983. Characterization of automotive emissions by bacterial mutagenesis bioassay: a re- view, Environ. Mutat. 5:609-631. Coordinating Research Council-Air Pollution Re- search Advisory Committee. 1985. Automotive benzene emissions, CRC-APRAC Project No. CAPE-35-83, Preliminary Report, October 1985, Atlanta, Ga. Eisenberg, W. C. 1983. Polycyclic aromatic hydrocar- bon round robin study, Final Report to the Coordi- nating Research Council, August 1983. Eisenberg, W. C., Schuetzle, D., and Williams, R. L. 1984. Cooperative evaluation of methods for the analysis of PAH in extracts from diesel particulate emissions, Society of Automotive Engineers Tech- nical Paper 840414, presented at the International Correspondence should be addressed to John H. John son, Michigan Technological University, Department of Mechanical Engineering and Engineering Mechan ics, Houghton, MI 49931. 73 Congress & Exposition, Detroit, Mich., February 27-March 4, 1984. Evans, S. B. 1980. Information report: 1979 progress  of the chemical characterization panel of the com- position of diesel exhaust project and results of particulate extraction roundrobin, Coordinating Re- search Council Report No. 516, Atlanta, Ga. Ford Motor Co. 1985a. Regulatory requirements and forecast tables, Dearborn, Mich. Ford Motor Co. 1985b. MVMA Vapor pressure sur- vey data, Dearborn, Mich. Fox, D. L. 1985. Air pollution, Anal. Chem. Appl. Rev. 57:22~238. Furey, R. L. 1985. Volatility characteristics of gaso- line-alcohol and gasoline ether fuel blends, Society of Automotive Engineers Paper 852116, Warren- dale, Pa. General Motors Corp. 1985a. Vehicle In-Use Emissions and Vehicle Emissions Modeling, Warren, Mich. General Motors Corp. 1985b. EPA/EMA in-use emis- sion factor test program, information available from Detroit Diesel Allison Division, Detroit, Mich. General Motors Corp. 1986. Pocket Reference. General Motors Corporation, Environmental Activities, April 1986, Warren, Mich. Gibson, T. L., Ricci, A. I., and Williams, R. L. 1980. Measurement of polynuclear aromatic hydrocar- bons, their derivatives, and their reactivity in diesel automobile exhaust, General Motors Corporation Research Publication 3478, ENV #5. Warren, Mich. Gorse, R. A., Jr., and Salmeen, I. T. 1982. Effects of filter loading and filter type on the mutagenicity and composition of diesel exhaust particulate extracts, Atmos. Environ . 16:152~1528. Gray, C. 1985. The EPA perspective on methanol as a motor vehicle fuel, Paper presented at EPA Region VI Methanol Workshop, May 20, 1985, Dallas, ,~ 1ex. Greco, R. 1985. Motor Vehicle Tampering Survey- 1984, U.S. Environmental Protection Agency, Of- fice of Air and Radiation, July 1984, Washington, D.C. Gross, G. P., MacDonald, J. S., and Shahed, S. M. 1982. Informational report on the measurement of diesel particulate emissions, Coordinating Research Council Report No. 522, Atlanta, Ga. Hampton, C. V., Pierson, W. R., Harvey, M. T., Updegrove, W. S., and Marano, R. S. 1982. Hy- drocarbon gases emitted from vehicles on the road. A qualitative gas chromatography/mass spectrom- etry survey, Environ. Sci. Technol. 16:287-298. Harvey, C. A., Carey, P. M., Somers, J. H., and Garbe, R. J. 1984. Toxicologically acceptable levels of methanol and formaldehyde emissions from methanol fueled vehicles, Society of Automotive Engineers Paper 841357, Warrendale, Pa. Heavenrich, R. M., Murrell, J. D., and Cheng, J. P. 1986. Light-duty automotive trends through 1986, Society of Automotive Engineers Paper 860366, Warrendale, Pa. Heinen, C. M. 1980. We've done the job-What's it worth? Society of Automotive Engineers Paper 801357, Warrendale, Pa. Herr, J. D., Dukovich, M., Lestz, S. S., Yergey,

74 J. A., Risby, T. H., and Tejada, S. B. 1982. The role of nitrogen in the observed direct microbial mutagenic activity for diesel engine combustion in a single-cylinder DI engine, Society of Automotive Engineers Paper 820467, Warrendale, Pa. Hilliard, J. C., and Springer, G. S. 1984. Fuel Economy in Road Vehicles Powered by Spark Ignition Engines, Plenum Press, London. Hyde, J. D., Gibbs, R. E. Whitby, R. A., Byer, S. M., Hill, B. J., Hoffman, T. D., Johnson, R. E., and Wenner, P. C. 1982. Analysis of particulate and gaseous emissions data from in-use diesel passenger cars, Society of Automotive Engineers Paper 820772, Warrendale, Pa. Jones, A. D., Appleby, M. R., and Bintz, L. J. 1982. Effects and detection of implanted defects on new technology emission control systems, Society of Automotive Engineers Paper 820977, Warrendale, Pa. Kimbara, Y., Shinody, K., Koide, H., and Koba- yashi, N. 1985. NOR reduction is compatible with fuel economy through Toyota's lean combustion system, Society of Automotive Engineers Paper 851210, Warrendale, Pa. Krost, K. J., Pellizzari, E. D., Walburn, S. E., and Hubbard, S. A. 1982. Collection and analysis of hazardous organic emissions, Anal. Chew. 54:810- 817. Lee, F. S. C., Pierson, W. R., and Ezike, J. 1980. The problem of PAH degradation during filter collec- tion of airborne particulates an evaluation of sev- eral commonly used filter media. In: Proceedings of the Fourth International Symposium on Polyaro- matic Hydrocarbons: Chemistry and Biological Ef- fects (A. Bjoerseth and A. J. Bennis, eds.), pp. 543-563, Columbus, Ohio: Battelle Press. Lipari, F., and Keski-Hynnila, D. 1985. Aldehyde and unburned fuel emissions from methanol-fueled heavy-duty diesel engines, General Motors Research Publication 5173, July 15, 1985, Warren, Mich. McNutt, B. D., Elliot, D., and Dulla, R. 1984. Patterns of vehicle misfueling in 1981 and 1982: where, when, what vehicles, and how often? Soci- ety of Automotive Engineers Paper 84135, Warren- dale, Pa. Motor Vehicle Manufacturers' Association. 1985. MVMA newsletter vehicle tampering, J. Air Pol- lut. Control Assoc. 35:1-8. Mulawa, P. A., and Cadle, S. H. 1979. Low-molec- ular weight aliphatic amines in exhaust from cata- lyst-equipped cars, General Motors Research Pub- lication 2946 ENV-57, Warren, Mich. National Research Council. 1983a. Feasibility and As- sessment of Health Risks fi~om Vapor Phase Organic Chemicals in Gasoline and Diesel Exhaust, National Academy Press, Washington, D.C. National Research Council. 1983b. Polycyclic Aro- matic Hydrocarbons: Evaluation of Sources and Effects, National Academy Press, Washington, D.C. Nichols, R. J., and Norbeck, J. M. 1985. Assessment of emissions from methanol-fueled vehicles: impli- cations for ozone air quality, Ford Motor Co., Dearborn, Mich.  Automotive Emissions Okamoto, W. K., Gorse, R. A., and Pierson, W. R. 1983. Nitric acid in diesel exhaust, J. Air Pollut. Control Assoc. 33:109~1100. Perez, J. M., Hills, F. J., Schuetzle, D., and Williams, R. L. 1980. Informational report on the measure- ment and characterization of diesel exhaust emis- sions, Coordinating Research Council-Air Pollu- tion Research Advisory Committee Project No. CAPI-1-64, CRC Report No. 517, Atlanta, Ga. Perez, J. M., Lipari, F., and Seizinger, D. E. 1984. Cooperative development of analytical methods for diesel emissions and particulates solvent extract- ables, aldehydes and sulfate methods, Society of Automotive Engineers Paper 840413, Warrendale, Pa. Pierson, W. R., Brachaczek, W. W., Korniski, T. J., Truex, T. J., and Butler, J. W. 1980. Artifact formation of sulfate, nitrate, and hydrogen ion on backup filters: Allegheny Mountain experiment, J. Air Pollut. Control Assoc. 30:30-34. Pitts, J. N., Jr., Biermann, H. W., Winer, A. M., and Tuazon, E. C. 1984. Spectroscopic identification and measurement of gaseous nitrous acid in dilute auto exhaust, Atmos. Environ. 18:847-854. Risby, T. H., and Lestz, S. S. 1983. Is the direct mutagenic activity of diesel particulate matter a sampling artifact? E'?viron. Sci. Technol. 17:621-624. Schuetzle, D., Lee, F. S-C., Prater, T. J., and Tejada, S. B. 1980. The identification of polynuclear aro- matic hydrocarbon derivatives in mutagenic frac- tions of diesel particulate extracts, Paper presented at the 10th Annual Symposium on the Analytical Chemistry of Pollutants, Dortmund, Germany, May 28-30, 1980. Schwarz, M. J. 1985. Onboard refueling control not a simple task, Paper presented at 78th Annual Meeting of Air Pollution Control Association, July 1985. Simon, G. M., and Stark, T. L. 1985. Diesel particu- late trap regeneration using ceramic wall flow traps, fuel additives, and supplemental electrical igniters, Society of Automotive Engineers Paper 850016, Warrendale, Pa. Sobey, A. J. 1985. GM's viewpoint on alternative fuels, presentation at joint U.S./Canadian Meeting on Alternative Fuel, Windsor, Canada, June 2026, 1985. Society of Automotive Engineers and U. S. Depart- ment of Energy. 1985. Conference and workshop on transportation fuel alternatives for North Amer- ica into the 21st century, Washington, D.C., No- vember 13-15, 1985, Society of Automotive Engi- neers, Paper 190, Warrendale, Pa. Stebar, R. F., Benson, J. D., Sapre, A. R., Brinkman, N. D., Dunker, A. M., Schwing, R. C., and Martens, S. W. 1985. Gasoline vapor pressure re- duction an option for cleaner air, Society of Au- tomotive Engineers, Paper 852132, Warrendale, Pa. Survey Data Research. 1985. MVMA Vehicle Tam- pering Study. Final Report, July 1985. Tupa, R. C., and Doren, C. J. 1984. Gasoline and diesel fuel additives for performance, distribution, quality, Society of Automotive Engineers Paper 841211, Warrendale, Pa. U.S. Department of Energy. 1985. The Motor Fuel

John H. Johnson 75 Consumption Model, 12th Periodic Report, U.S. DOE PE/77000-1, Washington, D.C. U.S. Environmental Protection Agency. 1972. Con- trol of air pollution from new motor vehicles and new motor vehicle engines. Fed. Regist. 37:2425 24337. U.S. Environmental Protection Agency. 1983. Anti- tampering and anti-misfueling programs to reduce in-use emissions from motor vehicles, EPA-AA- TTS-83-10, Research Triangle Park, N.C. U.S. Environmental Protection Agency. 1985. Com- pilation of air pollution emissionfactors) Vol. II, Mobile Sources, Washington, D.C. Wade, W. R., and Jones, C. M. 1984. Current and future light duty diesel engines and their fuels, Society of Automotive Engineers Paper 840105, Warrendale, Pa. Way, G. 1985. 1986 Engines and Emission Systems Tables (unpublished; Tables can be obtained by correspondence to Gilbert Way, 2371 Dorchester North, Troy, MI 48084). Weaver, C. S., Miller, C., Johnson, W. A., and Higgings, T. S. 1986. Reducing the sulfur and aromatic content of diesel fuel: costs, benefits, and effectiveness for emissions control, Society of Au- tomotive Engineers Paper 860622, Warrendale, Pa. Wiedemann, B., and Neumann, K. H. 1985. Vehicu- lar experience with additives for regeneration of ceramic diesel filters, Society of Automotive Engi- neers Paper 850017, Warrendale, Pa. Williams, R. 1985. Emissions measurement and unreg- ulated emissions, Ford Motor Co., Dearborn, Mich. Wineland, R. J. 1985. The Ford flexible fuel vehicle. Ford Motor Co., Dearborn, Mich.