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

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

Chapter 5 CONTROL COSTS AND OTHER REGULATORY QUESTIONS This chapter presents some regulatory issues that arise from the infor- mation on technology and health effects developed in previous chapters. The costs associated with this regulation will be discussed first. The conclusions of this chapter will focus on questions that need to be answered as part of any legislative or regulatory consideration of NOX emission controls on heavy-duty engines. CONTROL COSTS In a questionnaire, the committee asked engine manufacturers to estimate the costs imposed by tighter NOX standards in three categories: (l) hardware and initial testing, (2) any increased maintenance, and (3) the increased fuel consumption associated with emission control. The estimates of capital costs were very tentative, especially for heavy- duty gasoline engines. No information was received on maintenance costs, but these are expected to be small compared to the other cost categories. It became obvious, however, that any significant increase in fuel consumption would have costs greatly outweighing the initial capital costs. Thus, the consideration of the costs of NOX regulations must focus on the change in fuel consumption. Table 22 shows the information developed by EPA in its regulatory analysis. These numbers were based on particular assumptions about fuel economy and annual mileage, but other reasonable assumptions would show the same general relation between initial cost and fuel penalty. For diesel engines the fuel cost of a l-percent increase in fuel consumption is roughly equal to the initial capital cost increase. For gasoline engines, a 2-percent increase in fuel con- sumption gives a fuel cost increase approximately equal to the initial capital cost. In the remainder of this section we will consider only fuel costs. It is important to emphasize that the costs given in this section are illustrative only. Although their approximate magnitudes are known, and the conclusions we draw from them sound, actual fleet average data on fuel penalties and capital costs are necessary for accurate cost-effectiveness calculations. We have not evaluated, even in a qualitative fashion, the total benefits of NOX control. Nor have we provided any kind of cost-benefit analysis. These should be considered parts of a full regulatory analysis. 79

80 TABLE 22 EPA Cost Estimates for NOX Controls Cost Category Cost (Dollars) Gasoline Engines3 Diesel Engines^ Capital costc 270 733 Cost of l-percent increase" in fuel consumption l50 754 aControl technology: three-way catalysts substituted for oxidation catalysts. ^Control technology: turbocharging, charge cooling, electronic injection timing controls, and exhaust gas recirculation. GUndiscounted costs for hardware, R&D, and certification testing. "Lifetime fuel costs, undiscounted. SOURCE: Adapted from U.S. Environmental Protection Agency (1980).

8l We have evaluated some typical data on heavy-duty truck use to estimate the fuel consumption costs for different vehicle weight classes (Table 23). Spec- ifically, the costs per mile of a l-percent increase in fuel consumption ranges from 0.l2 cents (for diesels in weight classes 3-5) to 0.40 cents (for gasoline engines in classes 7-8). The same l-percent increase in fuel consumption increases the lifetime fuel cost by $258 to $l,l5l. Of course, the costs of a l-percent fuel penalty can simply be multiplied by the actual percentage increase in fuel consumption to obtain the actual costs. The implications of increased fuel consumption on the entire heavy-duty truck fleet are illustrated in Table 24. The projected fuel use is for l995, when most of the fleet miles will be driven by trucks produced in the mid-l980s and later. The data are from the base case in a paper by Jambekar and Johnson (l98l). The fuel use for class 2B (8,50l-l0,000 pounds gross vehicle weight rating) was estimated as 5.5 percent of the total light truck fuel use. The figure of 5.5 percent was taken from data in the EPA regulatory analysis (U.S. Environmental Protection Agency, l980), as the percentage of light trucks in class B. Table 24 shows that a l-percent increase in fuel consumption for the heavy-duty fleet leads to a total annual cost of $430 million per year; $397 million of this is for diesel fuel, of which $338 million is for class 7 and 8 trucks. From a regulatory perspective, one important number is the cost- effectiveness of a given degree of control (that is, the dollar cost divided by the emissions reduction). A proper cost-effectiveness accounting must consider the sales-weighted increase in fuel consumption for all vehicles in the fleet. This requires a detailed analysis that is beyond the scope of the committee's charge. To illustrate the relative cost-effectiveness of various controlled emissions levels for diesel and gasoline engines we have calculated the increase in fuel costs for assumed fuel penalties and assumed emissions levels. These cost-effectiveness numbers (for fuel costs only) are shown in Table 25. Although the percentage fuel penalty, at a given NOX level, is likely to be different for the two different engine types, Table 25 shows that the cost-effectiveness is better for the diesel when the penalty is the same for both engines. The cost-effectiveness for both engines is the same only when the percentage fuel penalty of the diesel is greater than that of the gasoline engine. There are two reason for this. First, control to a given level implies greater emissions reductions for diesel engines than for gasoline engines. Second, for a given percentage increase in fuel consumption, the more efficient diesel engine increases its actual fuel consumption less than the gasoline engine. INDIRECT COSTS OF NOV CONTROL A In assessing the cost of an NOX control regulation, it is important to con- sider increases in any other pollutant species cause by the control of NOX. The diesel engine chapter describes the increases in particulate and hydrocarbon emissions associated with increases in the stringency of NOx emission control.

82 TABLE 23 Fuel Costs for Typical Heavy-Duty Vehicles Data by Weight Class Classes 3-5 Class 6 Classes 7&8 Gasoline or diesel G D G D G D Assumed lifetime (years) 89 89 8 9 Lifetime mileage3 (thousands of miles) 127 247 126 247 137 503 Fuel economy in 1986b 6.84 10.88 6.03 9.77 3.49 5.90 Lifetime fuel use (thousands of gallons) 18.4 22.7 20.9 25.3 39.3 85.3 Assumed fuel cost (dollars per gallon) 1.40 1.35 1.40 1.35 1.40 1.35 Fuel cost/distance traveled (cents per mile) 20.5 12.4 23.2 13.8 40.1 22.9 Lifetime fuel cost (thousands of dollars) 25.8 30.6 29.3 34.2 55.0 115.1 Cost of a 1% Increase in Fuel Consumption (cents per mile) 0.205 0.124 0.232 0.138 (lifetime cost, in 258 '306 293 342 dollars) 0.401 0.22' 550 11! aData from Energy and Environmental Analysis, Inc. (1980). bData from Jambekar and Johnson (1981), Table 9.

83 H 01 rH 00 • -* r» 01 4J • • r^ •H O CT* ON ON P H CM co co oo 00 •o 00 a oo i Cd oo mm co CJ ^ CM PO CO ^*\ £ 00 rH 00 •1 • • iH 4J CO vO n in m O0 01 0 C o •H bo rH C m CM f> CO U • • • g r-> o o co § iH 01 4Q CO N.X 01 Cd ^r in •H CO • • m a CM o o m ON rH a C o iH 4-1 c rH CO 01 •H too 01 rH rH iH o cd ir» fa CO 4J ^3- CO CO CO o • • CO ,_j • ^» o H CM m en 01 i—j £ 3 CO 9 IH rH 01 M •^ CJ CO ^ 3 0 Cd •H 01 oo ^J T5 DM CO CO *o ^ • •t c cd c JJ a o rH CO rH rH rH 3 o co CJ ^ O O T rH 01 ^ 4-1 ^% CO •H Jo J= iH bO 41 bO 00 rH M CTN cO iH O cd CO • * CO rH rH ^f c 01 ^o O iH iH v_x • 01 a i_i iH Pu •H CO a bO 4J o iH C O CO 0} w 4 J c G ^ lO CO -tf -C •H PM (U I • • I4-1 o rH c o o •* O O M-4 •H CO o rH 4-1 *o cO O X"N a S bo 4-1 CO CO V cd o i Cd rH m r< o (4 •• o • • m cd M (O a cs rH rH rH rH 01 jj 4-1 g f^ a. co M O 01 O •o m ja o 4-1 • e CO C MH m 3 _— j O 0 0 , •-j 0 Q_3 •H O 00 9 rHtC s~^ CO rH '-^ 4J G CO cd C cd g •^ IHCO COOtl rH>>O 13 rl •O CM -DC O-HCO 14-1 0) 4-1 -H 0> *^ O OrHrH O3rHrH e g W CO rH rH rH 4H n) rH 3 cd 3 00 iHrH rHfHO 4-1 C -H 00 4-1 co m CO 01 td a1 -£> T3 COCCUS rH 3 bO 3 N-' O rH Cu^x CJ fa U. CJ CO CO cd CB o H a

84 TABLE 25 Fuel Costs at Selected Assumed NOX Emission Rates and Fuel Consumption Penalties3 Fuel Consumpt ion Penalty, (percent) Cost (dollars) per Ton of NOX Abated at NOX Emission Rates (g/bhp-h) of: Engine Type 765 4 3 2 Gasoline 1 1,400 700 470 350 2 2,800 1,400 930 700 s 7,000 3,500 2,300 1,750 10 14 ,000 7,000 4,700 3,500 20 28,000 14,000 9,300 7,000 Diesel 1 900 450 300 225 180 150 2 1,800 900 600 450 360 300 5 4,500 2,250 1,500 1,125 900 750 10 9,000 4,500 3,000 2,250 1,800 1,500 20 18,000 9,000 6,000 4,500 3,600 3,000 aThe assumed data are as follows : o Current brake-specific fuel consumption (BSFC) 200 g/bhp-h for diesel engines, 300 g/bhp-h for gasoline engines o Current NOX emissions (Ej) 8 g/bhp-h for diesel engines, 6 g/bhp-h for gasoline engines o Fuel costs (Cf) $1.35 for diesel fuel, $1.40 for gasoline o Fuel density (Pf) 6 Ib per gallon. The cost-effectiveness (CE) for a given controlled emission level £2 and percentage fuel penalty P is given by the equation CE- _ (P/100) (BSFC) (CfPf) E! - E2 With the above data and the necessary unit conversion factors, this equation yields the following computational equations: CE - 1400 P 6 - E2 for gasoline engines, and CE - 900 P 8 - E2 for diesel engines. Units are dollars per ton of NOX emissions abated.

85 Emissions of particulates, hydrocarbons, and carbon monoxide are governed by existing or proposed regulations; any NOX control technique that would increase these emissions above the regulated level would therefore necessitate additional control technology, and thus additional costs. Emissions of unregulated pollutants (e.g., odor or sulfates), which might also increase as a result of more stringent NOX control, could also be considered as imposing a cost in the form of additional health or welfare effects on the public. The committee has no quantitative information on the costs that can be attributed to emissions increases resulting from NOX control techniques. It is often difficult to determine how much of the emission control cost can be attributed directly to a specific pollutant species when the engine has been designed as a total package, which must meet all applicable standards. The capital costs of NOX control are difficult to assess. Particular engine design changes, for example, may improve performance and reduce fuel consumption in addition to controlling emissions; in such cases there is no simple way to say how much of the costs should be attributed to emission control. (Certification testing adds an additional capital cost, which is small compared with other cost elements.) Despite this uncertainty, which gives a range of $700 to $2,000 as estimates for the capital costs, per engine, of NOX controls, the easily calculated costs of any increase in fuel consumption due to NOX control become the largest component of the control costs if the corresponding increase in fuel consumption is greater than about 2 percent. REGULATORY ISSUES This section is not intended to be a full regulatory analysis. The questions discussed here arose in the course of the committee's study, but answering them will require work by the appropriate regulatory or legislative bodies. In addition, other important issues (e.g., the financial health of companies affected by the regulations) are not discussed here, but should be considered in developing regulations. Differences Between Gasoline and Diesel Engines The differences between gasoline and diesel engines have been described in previous chapters. From an emission control standpoint, the diesel engine has lower emissions of hydrocarbons and carbon monoxide, while the gasoline engine has lower NOX and particulate emissions. Catalytic emission control systems are widely used on light-duty gasoline engines, but their durability under the more severe operating conditions of heavy-duty engines remains to be proven. A catalytic system cannot be used on a diesel engine unless a supply of a reducing agent such as ammonia or methane is carried on board the vehicle to react with NOX. As noted in the introduction to this chapter, homogeneous reactions for NOX removal are not possible at the

86 temperatures and compositions of diesel exhaust. Diesel engines will find it harder than gasoline engines to achieve low NOX emissions without a large fuel penalty. On the other hand, because diesel engines are the largest fuel users of the heavy-duty engine fleet, any emission standard that imposes a fuel consumption penalty will have a greater effect on diesel users than on gasoline engine users. From a regulatory perspective, the cost of the increased fuel use required to reduce a unit amount of NOX emissions from diesels appears to be less than the comparable cost for gasoline engines. Given the differences between gasoline and diesel engines, it is appropriate to consider separate emission standards for the two engine types. In considering this question, it is necessary to analyze the relative competitive advantages of the two engines. There is already a trend toward the use of diesel engines in the lower weight ranges of the heavy-duty vehicle classes. This trend could be accelerated if separate standards improved the competitive advantage of the diesel (for example, by imposing heavier fuel consumption penalties on gasoline engines). The net impact of separate standards for diesel and gasoline engines, taking into account effects on emissions, the issue of relative competitive advantage, user costs, and the cost-effectiveness of emission controls, will be difficult to analyze. But such separate standards should be considered, if only because of the simple fact that it is easier to control the NOX emissions of gasoline engines than those of diesel engines. Vehicle Size Considerations The use of separate standards for the heaviest heavy-duty trucks (those in classes 7 and 8) has been suggested. As noted in the introduction, only l7 percent of the fuel used by these vehicles is used in urban areas, where emissions are most harmful. In addition, vehicles in these size ranges account for nearly three-fourths of the fuel used in heavy-duty vehicles. Thus, it appears that a higher NOx emission standard for these vehicles would have a small impact on urban air quality and would increase the total fuel consumption of the heavy-duty fleet less than a uniform regulation. Any analysis of separate standards for different vehicle sizes must consider the impact of their emissions on total air quality. The introductory section on emissions inventories reported a forecast of the California Air Resources Board that heavy-duty diesels (most of which are in classes 7 and 8) would contribute 24 percent of the total NOX emissions in the South Coast Air Basin in l987. This is only one area of the country, but it is the one with the highest annual average NOX concentrations. Any consideration of less stringent NOx standards for class 7 and 8 vehicles must examine the emissions that will not be reduced because of these less stringent standards, the impact of these foregone reductions on air quality, and alternative control strategies for obtaining the desired emission reductions.

87 Emissions Averaging "Emissions averaging" is a regulatory concept under which the emissions of a given population of engines are averaged for purposes of regulation, so that some may exceed the prescribed average emission rate so long as enough others have emissions lower than the average. There are several emissions averaging schemes, differing according to the populations over which the averages are taken. For example, the current system, under which each engine family from each manufacturer must meet the emissions standards on an average basis, determined by the regulations for acceptable quality limit (AQL), might be called emissions averaging, since a manufacturer can produce individual engines that do not meet the standards. However, what is generally meant by the term is averaging over a larger population, such as the entire output of an engine manufacturer, or even an industry. The concept of corporate averaging, which is currently used in the fuel economy regulations for passenger cars, has been suggested for emissions standards. The fuel economy regulations require that each manufacturer meet a corporate average fuel economy (CAFE) standard for its entire production. Under the corporate average emissions standard concept, a manufacturer could produce an engine family whose entire production exceeded the emissions standard provided that the same manufacturer also produced another engine family with emissions correspondingly below the standard. In concept, emissions averaging in heavy-duty engines could be applied as follows: Production of diesel engines would be used to meet the hydrocarbon and carbon monoxide standards and production of gasoline engines to meet the NOX and particulate standards, in recognition of the differences between the two engine types in terms of the difficulty of controlling emissions of each type. Production of the two engine types would be balanced to meet corporate average emission standards. If this were not possible, suitable controls could be placed on the engines whose emissions could be limited in the most cost-effective manner. The difficulty with this concept is that only two manufacturers (General Motors and International Harvester) produce both types of engines. EPA has suggested that manufacturers of different engine types could work cooperatively in reducing the emissions from their two (or more) companies to meet the average emissions standards. If this could be done, a form of the corporate averaging concept could be used. However, its use would be limited by the relative demands for the two types of engines. An alternative to averaging the emissions of the two engine types is the use of separate standards for gasoline and diesel engines, as discussed earlier. Another form of averaging is one that applies only to a single engine type. Here corporate average standards could be set for diesels and for gasoline engines; these standards could be the same or they could be different. Manufacturers could then determine which of their product lines could be controlled in the most cost-effective fashions. Since diesels would be averaged with diesels and gasoline engines with gasoline engines, this form

88 of averaging would not affect the relative competitive advantages of the two engine types. The U.S. Environmental Protection Agency (l98la) has announced that it will propose regulations which provide for averaging when it issues the final regulations for NOx emissions of heavy-duty vehicles, in May l982. The agency has not yet indicated which type of averaging it will propose. A fleet- average emissions standard could allow manufacturers to use emission-control technologies on only parts of their engine lines. This would provide in-use experience with new technology without the increased risk incurred when new technology is introduced over an entire product line. A fleet-average emissions standard that accomplished this could have a significant positive effect on the development of technology. Test Procedures Emissions reductions achieved in practice depend on the laboratory test procedures used to measure engine emissions. These test procedures should be reviewed continuously to ensure that emissions reductions measured in the laboratory are representative of actual, on-the-road emissions reductions from heavy-duty vehicles. The transient test cycle that EPA has mandated for heavy-duty emissions standards in l984 and subsequent model years represents a major change from the existing heavy—duty test cycle. (Appendix C describes and compares the old and new cycles.) In adopting the cycle (U.S. Environmental Protection Agency, l980a) EPA noted criticisms regarding the "justification for the tests, their repre- sentation of real life operation, their validation, their repeatability, and the lack of current knowledge upon which to base comments. After evaluating the comments EPA concluded that the test procedures were "necessary and appropriate." It further stated that: The origin of the operating cycles was an extensive program of actual in-use operational data collection and we are confident that the emphasis placed on quality during the subsequent cycle development program assures real world operating characteristics are well represented in the laboratory. It will not be possible to validate the cycle through time-consuming comparisons between on-the-road emissions and laboratory emissions. But, the pains taken to assure the representativeness of the cycles, in our view, would make such a validation superfluous anyway. EPA also concluded that all labs running the transient tests at the time the regulations were promulgated showed a "degree of correlation." The committee has received many criticisms of the cycle from engine manufacturers who disagree sharply with the conclusions of EPA. These disagreements are the subject of a current lawsuit by engine manufacturers

89 against EPA. This committee has not conducted a detailed analysis of the manufacturers' objections to the test cycle or of the EPA response to these objections; we offer no conclusions regarding the validity of arguments on either side of this issue. Our comments are directed to further considera- tions of the test procedures. EPA has been asked to determine whether the l984 heavy-duty truck requirements should be further revised based on the results of the manufacturers' current heavy-duty transient test programs (Executive Office of the President, l98l). Any test procedure for mobile source emissions represents a compromise between the need for an accurate representation of on-the-road emissions (which might dictate a long and complicated test procedure) and the need for a test procedure that can be easily and inexpensively performed by manufac- turers and control agencies. Any further evaluation of the test procedures by EPA should consider the appropriateness of test procedures with regard to this necessary compromise. It is also important to consider the accuracy and repeatibility of the test cycle. In a technical paper (Cox, l980), EPA has noted that "additional throttle calibration specifications for throttle controller performance are necessary to assure repeatibility" for the gasoline engine test procedure. The same paper concluded that "an additional tightening of the calibration 'window' for allowed torque variations appears necessary to assure repeati- bility." This indicates EPA's own recognition of the need for review of the test procedure. Table l2, in Chapter 2, noted the variation in emissions for the same engine when tested at different labs. It is important to have more cross-checks of this type to ensure the accuracy of the data used in any regulatory analysis. The ability of the test procedures to represent actual emissions per- formance should be subjected to an ongoing review. Regulations and Technological Feasibility Some recent environmental regulations have set standards, to take effect some time after promulgation, that could not be met by the existing technology. Such regulations (among them the proposed NOX emission regulations for heavy- duty vehicles), are called "technology forcing," and are often very contro- versial. Their use by legislative and regulatory bodies is based on the primacy of the health goal the standard seeks to achieve and the assumption that the regulation can prompt the development of the desired technology. There is often uncertainty about whether such standards can actually be met, and this uncertainty is of great concern to companies affected by the standard. The success of such regulations depends on a number of factors. The regulation should be perceived as firm by the industry being regulated. Regulations that are not firmly grounded technically, procedurally, or analytically may be perceived as likely to change; this perception impedes compliance efforts. Long-term consistency of approach on the part of the

90 regulators is especially important to firms with limited technical or financial resources, who may not be able to undertake costly development programs. It is also important that the regulatory requirements be consistent with the general momentum in the industry. For example, fuel economy standards have been achieved and surpassed because of the market forces that reinforced them. Conversely, regulations that contradict market trends or are so onerous as to force companies out of business are strongly resisted. Another factor in technology-forcing regulations is the possibility that additional companies may enter the emissions control business in response to the standard. Technology-forcing regulations must provide long lead times if they are to achieve significant technological change. The establishment of such regulations is a clear signal to other companies that a market is present if they can develop the necessary technology. In the specific subject considered here, the control of NOX emissions from heavy-duty engines, two clear problems would be raised by the establish- ment of a technology-forcing regulation: the durability of catalysts for control of NOX from gasoline engines and the effects of exhaust gas recir- culation on diesel engines' performance and durability. In each of these cases, technology development by industry would be influenced by the level of the standard that is ultimately set. This interaction between the setting of the standard and the technology that might become available to meet it has not been discussed by the committee. We have identified technologies likely to be available, based on an analysis of the current state of technology and of technology now under development. The question of what sort of technological developments will be promoted by what levels of standards is one that must be determined in the regulatory process. TECHNOLOGY BEYOND l986 This study has focused on heavy-duty emission control technology to be avail- able in time to meet the Clean Air Act's l986 deadline for promulgation of an NOX standard for heavy—duty vehicles. Chapters 2 and 3 note the pervasive uncertainty about future technology, an uncertainty taken account of in the Act's provisions for the setting of interim standards to be revised every three years. Because it is possible for the Administrator of the Environmental Protection Agency to set an interim NOX standard less stringent than the 75-percent reduction specified in the Act, and because the Act itself is currently undergoing review, it is important to consider technology that could become available over a longer period of time. The technologies discussed in this report will continue to evolve. Some technologies whose availability by l986 is uncertain will be available, for example, by l990. New engine designs, using improved technology, will still exhibit certain trade-offs, but they can provide improvements in all aspects

9l of engine operation—performance, emissions, and fuel economy—compared to current engines. Improved Fuel Controls Electronic fuel controls and improved fuel handling and injection systems for diesel engines are perhaps the most significant new technology likely to be widely available by the end of this decade. These systems are being developed now, and should be available on some production engines by the mid-l980s, yielding operational experience that will point the way toward improvements in design. Introduction of these systems over a period of years should allow manufacturers the time they need to alter their engine designs to take advantage of the new systems and their electronic controls. As discussed in Chapter 2, the main effect that improved fuel controls will have on NOX emissions will come from allowing the use of other control techniques (e.g. EGR) while minimizing the deleterious side effects encountered with current injection systems. Catalyst Technology The durability of catalysts for NOX emission control in heavy-duty gasoline engines has also been noted as a source of uncertainty. These catalysts have been shown to be effective when fresh. Additional tests of their durability will determine whether they can remain effective in actual heavy-duty engine operation. If not, it will be necessary to develop new catalyst formulations capable of sustaining the high exhaust temperatures encountered in heavy-duty gasoline engines. This testing and development should be completed in time for catalysts to be used on heavy-duty gasoline engines by l990. The use of catalysts for removing NOX from diesel engine exhaust is possible if a suitable reducing agent, such as ammonia, is added to the exhaust stream. One manufacturer who has studied such a system notes that a 50-percent reduction in engine-out NOX emissions can be thus obtained. Problems of catalyst durability, plugging, and catalytic performance over a sufficiently wide temperature range have been noted. This kind of-system is a longer term option, which may be available in the l990s. Ultimately, the use of such a system depends not only on its technological development, but also on the regulatory judgment that vehicle operators would faithfully refill the tank containing the reducing agent, when this action would have no perceptible effect on engine performance. Exhaust Gas Recirculation The extent to which exhaust gas recirculation (EGR) will be used in heavy-duty diesel engines by l986 is also uncertain. This technology too is undergoing

92 development and should be generally available by l990. One of the major problems with EGR is its likely effect on engine durability. The use of electronic controls to program EGR systems should improve this situation by allowing EGR rates to be varied so that the net effect is the greatest possible reduction in NOX emissions at the minimum effect on durability. Such programmed EGR systems should be available for widespread use by l990. Particulate Traps It has been observed in this report that with current control technology a decrease in NOX emissions from diesel engines generally implies an increase in emissions of particulates. However, it would be possible to reduce emissions of both types simultaneously if exhaust particulate traps were available. A number of firms are attempting to develop such traps, mainly for light-duty diesel engines (National Research Council, forthcoming). The Environmental Protection Agency (l98la) believes that these systems will be available for heavy-duty diesel engines in time for compliance with a l986 standard; this committee, however, is uncertain of this. Still, because of the variety of development programs in industry, it is quite possible that this technology will be available for use in l990 vehicles. (Although this technology would not directly reduce NOX emissions, it would allow the use of NOX control techniques that might otherwise be ruled out by their effects on particulate emissions.) Alternative Fuels and Engines New types of engines and fuels may be significant in the control of NOX emissions by the l990s, but it is clear that conventional engines and fuels will be used in almost all heavy-duty applications for at least the next decade. The use of alternative fuels in diesel engines is the subject of a recent Society of Automotive Engineers (l98l) compendium. Some commonly mentioned alternatives are alcohols and various emulsions (such as water or alcohols with diesel fuel). The use of such fuels will require not only the development of new engines that can burn them properly, but also the establishment of fuel manufacturing and marketing infrastructures. Despite these obstacles, the use of alternative fuels in diesel engines is receiving a great deal of attention because of the possibility that they can be burned with low particulate emissions (U.S. Environmental Protection Agency, l98lc). The status of these fuels for use in light-duty engines is reviewed in a National Research Council (forthcoming) report. The possibilities of lower emissions, as well as the attractions of potential alternatives to petroleum, should provide an impetus for further research in this area. The use of alternative engines is another possible means of reducing emissions over the long term. The forthcoming National Research Council report discusses alternative engine types in some detail. Although that report evaluates these engines for their applications in light-duty vehicles, its comments on the various engine types are generally valid for heavy-duty engines as well.

93 Two developments are of particular interest for heavy-duty vehicles. United Parcel Service (UPS) has contracted with Texaco and Ricardo Consulting Engineers to develop a stratified-charge version of their current General Motors gasoline engines, used in delivery vans.* Texaco has had this engine under development since the early l940s. If successful, the engine could replace the engines in about three-fourths of the current UPS fleet. UPS has chosen this approach because its experience with diesel engines in delivery vans has been unsatisfactory despite fuel savings as high as 20 percent. Previous versions of the Texaco stratified charge engine, in light-duty vehicles, have shown good fuel economy with moderate emissions, but stringent emission controls sharply increased its fuel consumption (Tierney et al., l975). The UPS sees its planned engine as a possible way to cut fuel consumption without the high initial cost of a diesel engine. Further work will show what kinds of emissions the engine can achieve. Another promising alternative engine for heavy—duty applications is the gas turbine being developed by a consortium of the Garrett Turbine Engine Company, the Mack Truck Company, and KHD (a German company).** This is a very large engine, rated at 550 horsepower in commercial applications. Its emissions of hydrocarbons, carbon monoxide, and nitrogen NOX have been measured at O.05, l.89, and 3.l3 g/bhp-h, respectively, on the steady-state test cycle. At its most efficient operating point it has a brake-specific fuel consumption of 0.393 Ib/bhp-h. Its particulate emissions have been reported as 0.33 grams per kilogram of fuel. Using the best fuel economy figure of 0.393 Ib/bhp-h gives a minimum brake-specific particulate emission rate of 0.38 g/bhp-h. (Of course, the actual emission rate over the test cycle would be greater than this, and no direct comparison with the proposed l986 particulate standard of 0.25 g/bhp-h, on the transient test procedure, is possible.) The developers of this engine are optimistic about its introduction in the latter half of the l980s, but they recognize that this high-powered engine's uses will be limited. Initial applications are expected to be in on-and-off-road applications such as logging and mining. The first application in trucks would be in class 8 trucks operating in rugged terrain. This particular engine is cited here because it is currently undergoing on-road evaluations in a truck. Future use of this engine will obviously depend on its ability to meet NOX and particulate standards, as well as on its fuel economy as compared to that of the diesel engines it would replace. Gas turbines and stratified charge engines are the most well developed of all alternative engines. Their ultimate use will depend on further develop- ment to reduce fuel consumption and emissions. The levels at which emissions standards are set in the future will obviously play a part in the development of these or any other alternative engine technologies. *Ruth A. Hunter, Massachusetts Department of Transportation, letter to Dennis Miller, January 20, l98l. **T. W. Qualle, Garrett Turbine Company, letter to Laurence S. Caretto, April 2l, l98l.

CONCLUSIONS Although data on the costs of NOX controls are preliminary and tentative, it is clear that any increase in fuel use of greater than about 2 percent, due to the imposition of NOX emission controls, will provide the greatest component of the control cost. This issue needs the most attention in determining the costs of compliance with an NOX control regulation. Issues that should be considered during the regulatory processes include: o The differences between gasoline and diesel engines o The different size ranges in heavy-duty engines o Emissions averaging concepts o The amount of research and development that would be promoted by a specific standard o The ability of the industry to respond to the regulation.

95 REFERENCES Cox, T. P. l980. "Heavy-Duty Gasoline Engine Emission Sensitivity to Variations in the l984 Federal Test Cyle." Warrendale, Pa.: Society of Automotive Engineers. (SAE Paper No. 80l370.) Energy and Environmental Analysis, Inc. l980. Medium- and Heavy-Duty Truck Fuel Demand Module Update and Calibration of the Highway Fuel Consumption Model. Arlington, Va.: Energy and Environmental Analysis, Inc.(U.S. Department of Energy contract no. DE-AC0l-79PE-70032, task 5) Executive Office of the President. l98l. "Actions To Help the U.S. Auto Industry." The White House, Office of the Press Secretary. April 6. Jambekar, A. B., and J. H. Johnson, l98l. "Effect of Truck Dieselization on Fuel Usage." Warrendale, Pa.: Society of Automotive Engineers. (SAE Paper No. 810022.) National Research Council. Forthcoming. Diesel Technology. Diesel Impacts Study Committee, Technology Panel. Washington, D.C.: National Academy Press. Society of Automotive Engineers. l98l. Alternate Fuels. Warrendale, Pa. (SP-480) Tierney, W. T., et al., l975. "The Texaco Controlled Combustion System, a Stratified Charge Concept: Review and Current Status." Paper presented at the Power Plants and Future Fuels Conference, Institution of Mechanical Engineers, January. U.S. Environmental Protection Agency. l980a. "Control of Air Pollution from New Motor Vehicles and Motor Vehicle Engines: Gaseous Emission Regulations for l984 and Later Model Year Heavy-Duty Engines." (Final rule.) Federal Register 45(l4):4l36. January 2l. . l980b. "Draft Regulatory Analysis, Environmental Impact Statement and NOX Pollutant Specific Study for Proposed Gaseous Emission Regulations for l985 and Later Model Year Heavy-Duty Engines." Office of Mobile Source Air Pollution Control. Washington, D.C.: U.S. Environmental Protection Agency. November 25. . l98la. "Control of Air Pollution from New Motor Vehicle Engines: Certification and Test Procedures." (Notice of intent.) Federal Register 46(70):2l628. April l3. . l98lb. "Control of Air Pollution from New Motor Vehicles and New Motor Vehicle Engines: Particulate Regulation for Heavy-Duty Diesel Engines." (Proposed rule.) Federal Register 46(4):l9l0. January 7.

96 l98lc. "Control of Air Pollution From New Motor Vehicle Engines: Gaseous Emission Regulations for l985 and Later Model Year Light-Duty Trucks and l986 and Later Model Year Heavy-Duty Engines." (Advance notice of proposed rulemaking.) Federal Register 46(l2):5838. January l9.