Breathing Clean - cti2000.it World Bank Methane buses.pdf · gas vehicle fleet in the world, has in fact no natural gas buses in regular operation. In devel-oping countries, diesel

WORLD BANK TECHNICAL PAPER NO. 516

Breathing CleanConsidering the Switchto Natural Gas Buses

Masami Kojima

The World Bank

Washington, D.C.

COPYRIGHT PAGE

iii

Contents

Foreword v

Abstract vi

Acknowledgments vii

Abbreviations and Acronyms viii

Executive Summary 1Natural Gas Vehicles: Economic and Technical Context 1Considerations for Policymakers 2A Social Choice 5

Chapter 1 Why Consider Natural Gas Vehicles? 7Natural Gas Vehicles: Some Basics 9Reasons for Switching Fuel to Natural Gas 13

Chapter 2 International Experience with Natural Gas Vehicles:Cases of Argentina and New Zealand 15

Argentina 15Background 15Lessons from Argentina 17

New Zealand 17Background 17Lessons from New Zealand 20

Other International Experiences 21Role of Government 21

Potential Government Assistance 21Designing Fuel Tax 22

iv Breathing Clean: Considering the Switch to Natural Gas Buses

Chapter 3 Comparison of Natural Gas and Diesel Buses 27Performance 27Emissions 29Fuel Quality 32International Experience 32

United States 33Australia 35Canada 36France 37Transit Bus Industry in Developing Countries 37

Chapter 4 Looking to the Future 41

Annex A: Emissions from Diesel Vehicles 45

References 49

BoxesBox 1 An Example of the Economics of CNG Buses in the United States 34Box 2 Phoenix Transit 36

TablesTable 1 Energy Content of Liquid and Gaseous Fuels 9Table 2 International Natural Gas Vehicle Statistics 12Table 3 Representative Fuel Prices 16Table 4 Percent Market Share of Liquid and Gaseous Fuels 16Table 5 Steps Taken by the Government of New Zealand 18Table 6 Emissions Benefits of Replacing Diesel with CNG Vehicles 29Table 7 Heavy-Duty Diesel Emission Standards

(in g/kWh, with g/bhp-h in parentheses) 30Table 8 Comparison of CNG and “Clean Diesel” Buses (g/km) 31

FiguresFigure 1 Payback for Conversion from Fuel Cost Savings in Months 17Figure A1 Particulate Emissions from New Vehicles in the United States 46

Foreword

T here is growing epidemiological evi-dence that emissions from conventionaldiesel vehicles are extremely harmful to

public health. Against this backdrop, natural gasbuses are attracting increasing attention in de-veloping country cities with serious air pollu-tion as policymakers explore alternatives toconventional diesel buses. A number of devel-oping country governments have announcedtheir intention to pursue the expansion of natu-ral gas bus fleets aggressively, including Chile,China, the Arab Republic of Egypt, India andIndonesia.

This report outlines technical, economic andpolicy issues that affect the success of naturalgas bus programs in developing countries. Theworldwide experience with natural gas buses islimited, but there are a number of importantlessons to be learned in those countries wherepolicymakers have attached a high priority tothe promotion of natural gas buses. We hopethat these lessons will be studied and integratedinto air quality management plans as govern-ments in developing countries consider the op-tion of switching from diesel to natural gas as afuel for buses.

Rashad KaldanyDirector

Oil, Gas and Chemicals Department

I n many large cities of the world, the trans-port sector is a significant contributor todeteriorating ambient air quality. One of

the most visible signs of urban air pollution isthe black smoke coming out of the tailpipes ofdiesel buses.

One option for effectively eliminating blacksmoke is to use natural gas instead of diesel. Asthis report shows, evaluating the costs and ben-efits of switching from diesel to natural gas foruse in buses raises a number of broader policyissues, ranging from inter-fuel taxation to restruc-turing of the transit bus industry. Merely man-dating natural gas buses, as some localgovernments have done without taking theseconsiderations into account, could endanger thesuccess of the natural gas bus program, seri-ously tarnishing its image in the eyes of not onlythe stakeholders in the energy and transportsectors, but also of the public.

We hope that this report will help stimulatefurther work and systematic data collection, andultimate assist policymakers to arrive at informeddecisions about the viability of natural gas busprograms in their cities.

Robert W. BaconManager, Policy Unit

Oil, Gas and Chemicals Department

v

I n response to emerging epidemiological evi-dence of the toxicity of diesel vehicularemissions, there is growing interest in sub-

stituting conventional diesel with much cleanernatural gas in cities where ambient concentra-tions of particulate matter are markedly higherthan what is internationally considered accept-able. This paper compares the performance ofnatural gas and conventional diesel buses, andoutlines the barriers to the adoption of naturalgas buses in developing countries.

In the absence of emissions standards thateffectively require natural gas, natural gas-fu-eled buses are unlikely to be adopted becausethey are more expensive to operate relative todiesel buses. This is partly because diesel is avery cheap fuel in most developing countries—it is lightly taxed or may even be subsidized.Even if diesel were sufficiently taxed, however,it is not obvious that natural gas buses would becheaper over their life cycle than diesel buses:they cost more to purchase, are less fuel effi-cient, and are often less reliable.

The above implies that the social case for re-placing diesel with natural gas as a fuel for busesrests on environmental grounds. If a local gov-ernment decides that the reduction in air pollu-

Abstract

tion associated with the substitution of conven-tional diesel with natural gas for use in buses isworth the cost, then it needs to adopt policiesto encourage the switch to natural gas. Thesepolicies might include emissions standards forbuses, or fuel and vehicle taxes that reflect mar-ginal social costs. In order to do so, the contri-bution of exhaust emissions from buses to theambient concentrations of harmful pollutantsneeds to be quantified so that associated healthdamage costs can be estimated—the benefits ofreducing emissions from buses must be higherthan incremental costs incurred. Further, success-ful implementation of fuel switching requires thata number of additional conditions be met: suffi-cient incentives for natural gas bus fleet opera-tors, regulatory and administrative arrangementsin place to ensure the financial sustainability oftransit operators who would be using naturalgas, large fleet operations converted to naturalgas to exploit economies of scale, proper regu-latory framework including enforced safety andperformance standards, strong and long-termcommitment and involvement of the fleet man-agement, extensive training and education ofmechanics and drivers, and regular preventivemaintenance and prompt repairs.

vi

M asami Kojima of the World Bank’s Oil,Gas and Chemicals Department (Glo-bal Products Group) prepared this

report under the guidance of Rashad Kaldany,Director of the Oil, Gas and Chemicals Depart-ment.

The sections in Chapter 2 on Argentina andNew Zealand are based on the presentationsmade by Juan Carlos Francchia, President of theArgentine Chamber for Natural Gas Vehicles, andGarth Harris, Secretary General of the Interna-tional Association for Natural Gas Vehicles, re-spectively, at the Workshop on Compressed

Acknowledgments

Natural Gas: An Option in Urban TransportProjects, held at the World Bank on 2-3 March2000.

The author thanks Robert Bacon of the Oil, Gasand Chemicals Department, Maureen Cropperof the Development Research Group, and KenGwilliam of the Urban Development and Trans-port Department, all of the World Bank, for theirconstructive comments; Paula Whitacre foreditorial assistance; and Annie Go Dizon fordesktop publishing. The conclusions and re-commendations of this report are solely thoseof the author.

vii

CARB California (Environmental Protection Agency) Air Resources BoardCBD central business districtCCC CNG Coordination Committee (in New Zealand)¢/l cents per literCNG compressed natural gasCPCB Central Pollution Control Board (of India)CRT Continuously Regenerating Technology (for particulate traps)EPA Environmental Protection Agency (in the United States)EPEFE European Programme on Emissions, Fuels and Engine TechnologiesEU European Uniong/bhp-h grams per brake horsepower per hourg/km grams per kilometerg/kWh grams per kilowatt-hourISO International Organization for Standardizationkg kilogramskm kilometerskWh kilowatt-hoursLNG liquefied natural gasm³ cubic metersmg/m³ milligrams per cubic meterMJ megajoulesMTA Metropolitan Transit AuthorityNG natural gasNGV natural gas vehiclesNOx oxides of nitrogenNPV net present valueNY New YorkNZ$ New Zealand dollarsOBD on-board diagnosticsOECD Organisation for Economic Co-operation and DevelopmentOEM original equipment manufacturersPM0.1 particles smaller than 0.1 microns

Abbreviations and Acronyms

viii

ix

PM2.5 particles smaller than 2.5 micronsPM10 particles smaller than 10 micronsrpm revolutions per minuteSAE Society of Automotive EngineersSCAQMD South Coast Air Quality Management District (of California)SOx oxides of sulfurUS$ United States dollarsUSAID United States Agency for International DevelopmentWHO World Health Organizationwt ppm parts per million by weightµg/m³ micrograms per cubic meter

Abbreviations and Acronyms

1

Executive Summary

E pidemiological evidence is emerging thatshows greater the toxicity of diesel ve-hicular emissions than previously be-

lieved. In response, there is growing interest insubstituting diesel with natural gas in cities whereambient concentrations of particulate matter aremuch higher than what is internationally con-sidered acceptable on health grounds. Are natu-ral gas vehicles an important component of thesolution? This report gives an overview of theissues that have to be considered when evaluat-ing natural gas as an alternative to diesel for usein transit buses.

Transit buses constitute one of the cheapestforms of mass transit. Traditionally fueled bydiesel, they are also significant emitters of fineparticles, which are known to cause prematuredeaths and illnesses. Urban transit buses are high-usage vehicles that operate in heavily congestedareas where air quality improvements and re-ductions in public exposure to harmful air con-taminants are critical. As such, they are goodcandidates for achieving emission reductions,and substituting natural gas for diesel is one wayof reducing emissions of fine particles and air-borne toxins dramatically. Particulate and air-borne toxin emissions from natural gas vehiclesare very low but not nil; the principal sourceof particulate emissions is the combustion oflubricant. Many natural gas buses in industrialcountries are equipped with an oxidation cata-lyst which reduces some of these emissionsfurther.

NATURAL GAS VEHICLES: ECONOMIC

AND TECHNICAL CONTEXT

Experience with natural gas buses is limitedworldwide. One reason is the relative inter-fueltaxation policies adopted by most governments,which make the end-user price of diesel lowerthan that of gasoline, and often more competi-tive than natural gas. As a result, the majority ofnatural gas vehicles operating in the world to-day are converted from existing gasoline ve-hicles. Argentina, which has the largest naturalgas vehicle fleet in the world, has in fact nonatural gas buses in regular operation. In devel-oping countries, diesel is seen as a social fueland is taxed little or even subsidized, making iteven more difficult to justify conversion to naturalgas on commercial grounds without large finan-cial incentives.

Environmental concerns drive the majority ofnatural gas bus programs today. Another impor-tant reason for turning to natural gas in trans-port is diversification of energy sources. Thisobjective is more easily achieved by targetinggasoline vehicles because of inter-fuel pricing.The United States is a world leader in deployingnatural gas-fueled transit buses in cities withserious air pollution. Some Asian countries, suchas China, India and the Republic of Korea, arealso aggressively pursuing natural gas bus pro-grams. Much technical progress has been madein the heavy-duty natural gas vehicle sector sincethe early 1990s. Consistent reports show that

2 Breathing Clean: Considering the Switch to Natural Gas Buses

natural gas buses manufactured in the early 1990shad not only a higher capital cost than their die-sel equivalents, but also were about 30 to 40percent more expensive to maintain and wereless reliable (for example, as measured in termsof the mean time between in-service failures,down to half that of diesel vehicles). The latestmodel heavy-duty natural gas vehicles are muchimproved, although the engine technology stillneeds some refinements.

Buses fueled by natural gas may be dual-fuel,running on diesel and natural gas with the com-bustion of diesel used to ignite the natural gas,or dedicated, running entirely on natural gas.Because of the stop-and-start nature of urbanbuses, the substitution of diesel with natural gasis limited in dual-fuel buses, and dedicatedsingle-fuel buses are recommended. Diesel en-gines, which are compression ignition based, aremore difficult to convert to dedicated naturalgas engines, which are spark ignition based, thangasoline engines. Thus, from the point of viewof emissions reduction as well as overall perfor-mance, engines and vehicles provided by origi-nal equipment manufacturers (OEM) aregenerally accepted to be superior, even if moreexpensive, than converted vehicles, and themajority of natural gas buses in industrial coun-tries are OEM vehicles. Natural gas buses arecleaner, quieter, and have less vibrations andodors than their conventional diesel equivalents.The fuel economy of natural gas buses is lowerthan that of diesel buses on an energy equiva-lent basis, by at least 10 to 15 percent and typi-cally even more. They have a shorter drivingrange, often less than two-thirds of diesel, sothat if refueling occurs only at depots, bus routesmay have to be managed differently. On thewhole the experience to date suggests that natu-ral gas vehicles are less reliable than diesel ve-hicles, although vehicle manufacturers areaddressing this. One of the important compo-nents of a successful natural gas vehicle pro-gram is extensive training of mechanics anddrivers, and the availability of qualified engi-

neers for technical support. Training is needednot only for proper maintenance and safeoperation of vehicles, but also to dispelmisperceptions and build the acceptance andcommitment of the operators involved. In Sydney,Australia, for example, drivers had perceptionsof lack of acceleration and poor drivability ofnatural gas vehicles compared to diesel. Com-parison trials with diesel buses showed that thedrivers were confusing lack of noise from natu-ral gas buses with lack of acceleration.

CONSIDERATIONS FOR POLICYMAKERS

Before embarking on natural gas bus programs,it is important to confirm, even if only order-of-magnitude figures are available, that contribu-tions from diesel vehicle emissions indeedconstitute a sizable fraction of ambient particu-late concentrations. Wrong assumptions aboutwhich sources are responsible for air pollutioncan lead to choices of measures that are notcost-effective or do not have a measurable im-pact on air quality. If the relative contribution ofroad traffic to pollution is actually small com-pared to other sources—such as refuse burning,emissions from cottage industries in the infor-mal sector, combustion of biomass in house-holds, and small diesel power generatorsoperated by shop owners placed on streets withmany pedestrians—then aggressively targetingvehicle exhaust for reduction will almost cer-tainly fail to improve air quality markedly. Whileit is difficult to identify sources accurately,chemical analysis of particles and other analyti-cal studies go a long way in providing a betterunderstanding of source contributions.

In cities where the contribution of transportto the ambient concentrations of fine particulatematter is deemed to be significant, replacingdiesel with natural gas in transit buses couldcontribute to a measurable improvement in airquality. In these cases, the following observa-tions from worldwide experience with natural

3Executive Summary

gas vehicles, and natural gas buses in particular,are worth considering.

Existence of natural gas pipeline. The volumeof gas consumed in the transport sector is notsufficient to justify the construction of naturalgas distribution pipelines even in large cities.Without the existence of a network of pipelinesfor other users of natural gas (industrial, com-mercial or domestic), a viable natural gas ve-hicle program would not be possible.

State of the transit bus industry. Transit buscompanies in many, if not most, developingcountries are cash-strapped. A large number ofoperators suffer from fare controls that havemade it very difficult to provide high-qualityservice. The emergence of mini-buses in the in-formal sector—that is, buses in the hands of non-corporate operators, illegal as well as legal—hasposed a serious threat to the survival of transitbuses, especially in the former Soviet Union andAfrica. Because traditional bus operators are of-ten cash-strapped, they do not maintain vehicles,nor can they purchase more expensive naturalgas buses, provide extensive staff training onthis new technology, and accept the possibilityof more repairs to deal with greater frequencyof bus breakdowns, at least initially. High emis-sions from diesel buses are not merely due tothe choice of fuel, but are often symptomatic ofdeeper problems in the transit bus industry indeveloping countries, and these same problemsmay condemn natural gas bus programs to fail-ure. If transit operators are in too weak a finan-cial position to switch to natural gas buseswithout some outside assistance, there shouldat the least be regulatory and administrative ar-rangements in place to ensure the operators’ fi-nancial sustainability when they use natural gas.

Inter-fuel and vehicle taxation policy favor-able to natural gas. Natural gas vehicles are moreexpensive than vehicles powered by liquid fu-els: vehicles are more expensive to purchase,

refueling stations may have to be established ata significant cost (bus fleet operators set up theirown refueling stations in industrial countries),and many natural gas bus operators have foundthat maintenance costs are also higher. For anatural gas vehicle program to be financiallysustainable in the long run, the incremental costmust be recovered in the form of fuel cost sav-ings, possibly supplemented by a large vehicletax difference in favor of natural gas vehicles.The viability of a natural gas vehicle programtherefore rests critically on the fuel and vehicletaxation policy adopted by the government, thefirst of which determines the relative prices offuels.

In industrial countries, tax is a large fractionof the final price of liquid fuels, so that it can beadjusted to favor one fuel over another. Becausetax collection in general is more efficient, differ-entiated vehicle taxes—whereby diesel vehiclesare taxed more than natural gas vehicles—present a possible option for favoring naturalgas. In most developing countries, gasoline car-ries a high tax rate, but diesel much less. If die-sel is taxed little or even subsidized, it may notbe possible to have an end-user price differ-ence between diesel and natural gas that is largeenough to achieve a reasonable payback periodwithout requiring other subsidies. Because taxcollection in general is less efficient, differenti-ated vehicle taxes may be difficult to implement.

A number of factors need to be taken intoaccount in designing fuel taxes, and they aredescribed in some detail at the end of Chapter2. Briefly,

� Diesel is an intermediate good, and under anarrow set of conditions, intermediate goodsshould not be taxed. These conditions, how-ever, are not met, especially in developingcountries.

� Goods that are close substitutes should carrycomparable tax rates. Gasoline and diesel (forlight-duty vehicles), natural gas and fuel oil(in industry and power generation), and natu-


ral gas used in transport on one hand and gaso-line and automotive diesel on the other areall close substitutes, certainly in the long run.

� Fuel and vehicle taxes should reflect envi-ronmental health risks caused by vehicles.Because diesel emissions are more harmfulthan gasoline, let alone natural gas, emissions,diesel vehicle owners operating in denselypopulated areas with serious air pollutionshould be asked to pay more to reflect theprice that society pays in the form of medicalcosts and productivity loss. However, fueltaxes that are typically set at the national leveldo not target negative environmental exter-nalities in urban areas well.

Government subsidies. In most countries withsuccessful natural gas bus programs, beginningwith the United States, the government has pro-vided significant subsidies. Subsidies weredeemed necessary because the inter-fuel pricedifferences were not sufficient to justify naturalgas programs purely on economic grounds, par-ticularly given that natural gas buses embody arelatively new technology with all the problemsthat accompany any new technology. However,unless government subsidies are consistentlymaintained, the threat of the suspension of gov-ernment subsidies discourages the growth of themarket. Providing heavy subsidies to “kick-start”the industry may also seriously distort the mar-ket, while the withdrawal of subsidies after afew years on account of “having given the in-dustry a chance,” could lead to the collapse ofthe market. This occurred in New Zealand, wherethe new Labor Government withdrew all gov-ernment support in the mid-1980s. In develop-ing countries, there are many competing andcompelling claims on the government budget,including provision of access to clean water,adequate health care, and universal primaryeducation. The relative merits of giving subsi-dies to the natural gas vehicle industry versusother social needs should be carefully con-sidered.

Regulatory framework. One of the most im-portant roles of the government is to establish aproper regulatory framework—in this case forthe natural gas industry and the transportsector—to eliminate market distortions as muchas possible, create a level playing field, ensuresafe operations, and increase efficiency andquality of service through competition. Interna-tionally acceptable standards for gas cylinders,refueling stations, gas dispensing units, conver-sion kits, natural gas vehicle and engine manu-facture, garages and the quality of gas shouldbe set. Equally important, an adequate monitor-ing and inspection system to enforce these stan-dards has to be in place. The same applies tothe transit bus industry. If poorly maintaineddiesel buses that are gross emitters are not forcedto comply with emission standards, or worse, ifthere are no emission standards to comply with,then the operating costs of diesel buses are ef-fectively lowered, making it difficult for cleanerbut more expensive natural gas buses to com-pete with diesel buses.

Economies of scale. In order for the operationto be financially viable, a large number of busesshould be made to run on natural gas, ideally awhole depot at a time. U.S. and French casestudies seem to suggest that a fleet size of tensof buses is desirable, preferably all located atone depot. This in turn may make conversionto natural gas difficult in countries with a largenumber of small operators, each owning two orthree buses.

Mandating natural gas / Emission standards.Emission standards can be made so stringentthat only gaseous fuel-powered vehicles, but notconventional diesels, can meet them. For heavy-duty vehicles such as urban transit buses, thiswould be tantamount to mandating natural gas.In the United States, the California Air ResourcesBoard (CARB) is developing very tight emissionstandards that conventional diesel cannot meet.In Delhi, India, the Supreme Court banned diesel

5Executive Summary

buses effective 2001. Such steps should be takenonly if certain conditions are met: overwhelm-ing evidence that diesel vehicles contribute sig-nificantly to ambient concentrations of particulatematter (and oxides of nitrogen, NOx, in citieswhere ozone is a serious problem and whereNOx reduction is believed to reduce ozone con-centrations); and the incremental cost of switch-ing to natural gas is greatly outweighed by thehealth benefits accruing from lower emissionsof natural gas buses. As mentioned earlier, otherimportant sources of fine particles, including theinformal sector and households, may also con-tribute considerably to worsening air quality indeveloping country cities.

Technology developments in industrial coun-tries. The greatest competitor to the natural gasvehicle industry in industrial countries today isperhaps the advent of the so-called clean dieselvehicle technology. Using ultra-low sulfur die-sel, these vehicles are equipped with catalyzedparticulate traps and other advanced controls.The future technology could include selectivecatalytic reduction or NOx adsorber traps for NOx

control. There are also advances made andbreakthroughs announced in refinery process-ing technology, potentially making the produc-tion of ultra-low sulfur diesel much cheaper inthe coming years. If these technologies currentlyunder development are successfully commercial-ized, the landscape for the clean diesel-naturalgas debate may dramatically change. At the sametime, the science of assessing the impact of fineparticles on public health—more specifically therole of particle size and chemical composition—and of measuring particles from vehicles and inthe atmosphere by number and size rather thanby mass (as is currently done) is rapidly evolv-ing. These developments will have a large im-pact on the future of the natural gas vehicleindustry in industrial countries, and will also af-fect the availability of natural gas vehicles indeveloping countries in the foreseeable future.

A SOCIAL CHOICE

To date, natural gas buses have been at a pri-vate economic disadvantage compared with die-sel buses unless supported by substantialfavorable tax discrimination or subsidies. In theabsence of emissions standards that effectivelyrequire gaseous fuels, natural gas buses are un-likely to be adopted because they are moreexpensive to operate relative to diesel buses.This is partly because diesel is a very cheapfuel in most developing countries—it is lightlytaxed or may even be subsidized. Even if die-sel were taxed much more, however, it is notobvious that natural buses would be cheaperover their life cycle than diesel buses: theycost more to purchase, are less fuel efficient,have a smaller range and are often less reliable.These observations suggest that the social casefor replacing diesel by natural gas in buses restson environmental grounds. In particular, the useof natural gas by heavy-duty vehicles normallyfueled by diesel would not be suitable if thediversification of energy sources is the primaryobjective.

Until such a time as clean diesel becomeswidely available on the international market,which is not expected for at least several moreyears, most developing country cities will haveto continue to grapple with a choice betweenconventional, polluting diesel versus potentiallymuch cleaner natural gas buses. If the govern-ment of a city decides that the reduction in airpollution associated with natural buses is worththe cost, then it needs to adopt policies thatwould encourage the switch to natural gas: ei-ther emissions standards for buses, or fuel orvehicle taxes that reflect marginal social costs.Once the decision to switch to natural gas hasbeen made, it is important to check if the condi-tions for successful implementation of fuelswitching are likely to be met: sufficient incen-tives for natural gas bus fleet operators, the regu-latory and administrative arrangements in place


to ensure the financial sustainability of transitoperators who would be using natural gas, largefleet operations converted to natural gas to ex-ploit economies of scale, proper regulatoryframework including enforced safety and per-

formance standards, strong and long-term com-mitment and involvement of the fleet manage-ment, extensive training and education ofmechanics and drivers, and regular preventivemaintenance and prompt repairs.

7

Chapter 1

Why Consider Natural Gas Vehicles?

Under the right set of circumstances, masstransit can offer greater sustainabilityand carrying capacity than private au-

tomobiles. Transit buses constitute one of thecheapest forms of mass transit. As such, busesare the backbone of the motorized transportsystem in most cities in developing countries.However, the sight of poorly maintained busesbelching out black smoke is all too common,tarnishing the image of public transport and pro-moting the perception that buses are the motor-ized transport mode of last resort. One of themost dramatic responses to the environmentalhealth impact of urban buses is the IndianSupreme Court’s total ban, effective 2001, ondiesel buses in Delhi. Even in the United States,the California Air Resources Board (CARB) hasrecently commented that current diesel busesusually emit more pollutants than if the bus rid-ers drove alone in their cars (CARB 1999).

As the Delhi case illustrates, diesel emissionsare under increasing attack based on emergingepidemiological evidence. CARB identified par-ticulate emissions from diesel-fueled engines astoxic air contaminants in August 1998 andlaunched a diesel risk reduction program. Thegovernments of industrial countries have re-sponded to the evidence by tightening dieselemission standards considerably. For example,the diesel emission standards to be implementedin phases in the United States beginning in 2006,and in the European Union (EU) beginning in2005 and further tightened in 2008, aim to re-

duce particulate emissions for new vehicles dra-matically, by ten-fold compared to the currenttechnology vehicles in the case of the UnitedStates. Annex A presents a more detailed dis-cussion of emissions from diesel vehicles andtheir effects.

In many developing country cities, the toxic-ity of diesel particulate emissions is of evengreater concern—diesel particulate emission lev-els are much higher than in industrial countrieson account of less advanced vehicle technol-ogy, poorer vehicle maintenance and poorer fuelquality, while the ambient concentrations of re-spirable particulate matter (PM10, particles smallerthan 10 microns) already far exceed internation-ally acceptable health-based standards. A largenumber of studies have linked exposure to el-evated levels of respirable particulate matter topremature deaths, hospital admissions and acuteand chronic illnesses. There is increasing evi-dence that the particle size also matters, withhealth effects worsening as the particle size de-creases. Because particles in vehicle exhaust arepredominantly in the small, sub-micron rangeand numerous, and they occur near ground levelwhere people live and work, they cause muchgreater human exposure in the immediate lo-cality than do emissions from sources such aspower plants for which stacks are situated atelevated levels and farther away from densepopulation centers. Equally disturbing, thethreshold level below which health effects arenot observable has not been identified, prompt-


ing governments in industrial countries to im-pose increasingly tighter standards—includingthe introduction of standards for PM2.5, particlessmaller than 2.5 microns—and the World HealthOrganization (WHO) to rescind its earlier health-based guideline values for particulate matter (onthe grounds that no safe threshold level has beendefined).

Particles emitted directly from a source aretermed primary; particles that are formed withinthe atmosphere, mainly from the chemical oxi-dation of atmospheric gases, are termed second-ary. Diesel is especially prone to high levels ofsmall primary particulate emissions because die-sel is heavier than gasoline and hence more dif-ficult to burn. The contribution of traffic to smallparticle emissions can be illustrated by takingan example from the United Kingdom: a recentstudy (Airborne Particles Expert Group 1999)concluded that in 1996, road traffic sources wereresponsible for only 25 percent of PM10 but for60 percent of PM0.1 (particles smaller than 0.1microns). Vehicle exhaust also includes twosources of secondary particles, oxides of sulfur(SOx) and of nitrogen (NOx). High levels of sul-fur in diesel (for example, close to 1 percent asfound in Jordan or Pakistan) would be expectedto contribute to significant secondary particleformation. NOx, the other pollutant of concernfound in diesel exhaust, similarly contributes tosecondary particles.

In contrast, the ambient concentrations ofother pollutants that are elevated in industrialcountry cities, such as ozone, are typically lowerin developing country cities. This is partly be-cause of lesser use of gasoline vehicles, withnotable exceptions such as Mexico City andSantiago de Chile that have serious ozone pol-lution. While there is a drive in industrial coun-tries to limit NOx emissions, an ozone precursor,to levels that can be achieved only by usingemerging technologies that are not yet commer-cialized, setting stringent NOx emission standardsis not expected to become a priority in the ma-jority of developing countries in the foreseeable

future. Instead, the primary focus will remainreducing particulate emissions, and the contri-bution of NOx to the ambient concentrations ofparticulate matter is relatively small.

Against this backdrop, natural gas (NG) hasbeen proposed as a much cleaner alternative toconventional diesel. Consisting primarily ofmethane and other light hydrocarbons, naturalgas does not contain hydrocarbons that formharmful emissions for the most part. In fact, theprincipal source of particulate emissions fromnatural gas vehicles (NGVs) is the combustionof lubricant. Many NGVs in commercial pro-duction already meet future particulate emis-sion specifications to be imposed in NorthAmerica and the EU during the latter half ofthis decade. Therefore, replacing heavy-dutydiesel vehicles with natural gas equivalents isone option for reducing vehicular particulateemissions dramatically.

Urban transit buses are high usage vehiclesthat operate in heavily congested areas whereair quality improvements and reductions inpublic exposure to harmful air contaminants arecritical. As such, they are good candidates forachieving both near-term and long-term emis-sion reductions. That many transit buses arecentrally kept and fueled makes the introduc-tion of new technologies and alternative fuelsmore efficient. In fact, natural gas vehicles areideal for fleet operations, and the natural gasindustry is concentrating on high fuel-use com-mercial vehicles such as transit buses, taxis, air-port shuttles, refuse haulers and trucks in itsmarket strategy.

This report gives an overview of the issuesthat have to be generally considered when evalu-ating natural gas as an alternative to diesel foruse in transit buses. This chapter gives back-ground information on NGVs and associatedinfrastructure. Chapter 2 gives a broad overviewof the development of the NGV industry in se-lect countries. Because the natural gas vehicleindustry worldwide consists mostly of light-dutygasoline vehicles converted to run on natural

9Why Consider Natural Gas Vehicles?

gas, the discussion focuses primarily on this cat-egory of vehicles. Chapter 3 then turns to natu-ral gas buses, highlighting advantages anddisadvantages over diesel, and drawing lessonsfrom international experience. Chapter 4 con-cludes with observations and a summary of is-sues to consider in evaluating the option ofpurchasing natural gas buses.

NATURAL GAS VEHICLES: SOME BASICS

Natural gas consists of the lightest hydrocarbons,inert gases (such as carbon dioxide), and negli-gible sulfur. The octane number of natural gascan exceed the scale’s maximum number of 120.To quantify the quality of the natural gas, meth-ane number is used as one measure. It is anexperimentally derived number correlating en-gine performance and fuel composition, depen-dent primarily on the content of methane andhigher hydrocarbons. Pure methane as the mostknock-resistant reference fuel is given a valueof 100. Similar to the octane number for gaso-line, a minimum methane number, which is afunction of engine technology, is needed to pre-vent engine knocking. The minimum methanenumbers for the current technology heavy-duty,advanced heavy-duty and light-duty vehicles areabout 80, 73 and 65, respectively (CARB 2001).

Another parameter that characterizes the en-gine behavior is the Wobbe index. The Wobbeindex, having the units of energy per unit vol-ume, is a comparative measure of thermal en-ergy flow through a given size orifice. If theWobbe index remains constant, changes in gascomposition will not lead to a change in air-to-fuel ratio and hence gases with the same Wobbeindex are interchangeable.

In terms of energy content, 1 kilogram (kg)of NG is equivalent to about 1.3 liters of gaso-line and 1.2 liters of diesel. On a volume basis,1 normal cubic meter (m³) of NG is equivalentto about 1.1 liters of gasoline and 1.0 liter ofdiesel. The relative energy efficiencies of en-

gines have to be factored into these figures toarrive at vehicle fuel economy. In order to storesufficient natural gas on board a vehicle toachieve an adequate driving range, natural gasmust be stored in high pressure tanks as com-pressed natural gas (CNG) or as cryogenic liq-uefied natural gas (LNG) in a highly insulateddewar. The volumetric energy content of thevarious fuels as stored, expressed in megajoules(MJ) per liter, is shown in Table 1. The advan-tage of liquid fuels is clear. In the United States,the fuel tank volume of CNG and LNG busesare about five times and twice that of diesel (Watt2001).

Natural gas as a transport fuel has a numberof advantages over diesel:

� Very low particulate emissions� Low emissions of airborne toxins� Negligible SO

x emissions

� More quiet operation, with less vibrations andless odors than the equivalent diesel engines.

All of these benefits make NGVs especiallysuitable in urban areas. In addition, life cycleanalysis suggests greenhouse gas emission sav-ings relative to gasoline, and possibly small sav-ings relative to diesel.

The disadvantages of natural gas include thefollowing:

� Greater difficulty in distribution and storage� Shorter driving range� Greater weight of the fuel tank (gas cylinder)

Table 1. Energy Content of Liquidand Gaseous Fuels

Relative RelativeFuel MJ/liter to gasoline to diesel

Gasoline 32 1.0 0.9Diesel 35 1.1 1.0CNGa 10 0.3 0.3LNG 19 0.6 0.5

a. CNG stored at 200 bar.Source: Maxwell and Jones 1995.


� Longer refueling time, especially if using aslow fill refueling system

� Backfire in the inlet manifold (which occurswhen hot gas from the cylinder escapes intothe inlet manifold and ignites the mixture).

Natural gas can be distributed economicallyin a city only by pipeline. Further, the amount ofnatural gas used in transport is not sufficient tojustify the construction of a pipeline networkeven in large cities, so that unless pipelines arealready in place or are planned for other uses ofnatural gas (such as for industrial and domesticpurposes), NGV programs are not viable. Ballparkfigures help to illustrate this point. In the year 2000,CNG and LNG vehicles consumed 385 million m³of natural gas in total in the United States. Incomparison, a 500-megawatt power plant oper-ating 5,000 hours a year (57 percent utilization)at 50 percent efficiency consumes about 500million m³ of gas. That is to say, one 500-mega-watt power plant consumes more natural gas thanall NGVs in the United States put together.

Because the price of natural gas has to bevery low for it to be competitive with liquid fu-els (see below), a cheap source of natural gas isneeded. A country with domestic production ofnatural gas is a much more likely candidate fora natural gas vehicle market than a gas-import-ing country. Storage of natural gas on board avehicle is costly because it can be stored only asCNG at about 200 bar (200 times the atmosphericpressure) and ambient temperature, or as a liq-uid (LNG) at –162°C at 2 to 6 bar. LNG vehiclesare much less common than CNG vehicles.

In order to store a reasonable amount of gasat 200 bar, large fuel tanks with thick walls areneeded, resulting in extra weight added to thevehicle. The use of composite materials can re-duce the tank weight considerably, but at ahigher cost. One area of research is the storageof natural gas at relatively low pressures bymeans of adsorption of hydrocarbon moleculesonto a structure with a large surface area, suchas activated carbon. The extra weight of the gas

cylinders currently used commercially increasesthe fuel consumption of NGVs, and potentiallyaccelerates the tire and brake wear. The extraspace taken by the fuel tank is a concern espe-cially in smaller vehicles such as taxis (in whichthe trunk space is reduced), but much less sofor larger vehicles such as transit buses, althoughthe extra weight reduces the passenger carryingcapacity.

During refueling, natural gas has to be com-pressed to a pressure in the neighborhood of200 bar, typically requiring about 0.2-0.3 kilo-watt hours (kWh) of energy per cubic meter ofgas. Refueling is one of the least safe momentsin the use of natural gas as a transport fuel. In arecent example from Delhi, a car converted torun on CNG exploded during refueling as thegas cylinder failed, injuring five people (Auto-motive Environment Analyst 2001). The causewas quickly identified to be the poor quality ofthe gas cylinder.

There are two types of NG refueling systems:fast fill and slow fill. A fast fill takes only a fewminutes. A slow fill costs less to set up but takeshalf an hour or more to fill a tank. However, aslow fill can be carried out at night when ve-hicles are not being operated, and gets moregas into the tank than a fast fill. In a slow fill, asecond short filling can be done easily to com-plete the first filling. In Poitiers, France, fast fill-ing was found to result in 15 to 20 percentunder-fill. Any under-filling reduces the drivingrange of the vehicle further. In the United States,typical costs for establishing a refueling stationfor 200 buses is of the order of US$0.35 millionfor diesel, US$0.95 million for LNG and US$2.7million for CNG (Watt 2001).

Because natural gas is lighter than air, it willnot lie along the ground if it leaks, and is thussafer in an accident. LNG, on the other hand,forms a liquid pool when spilt. Large accumula-tions of natural gas vapor can occur, resulting infire or explosion if an ignition source is nearby.Parking CNG vehicles in an enclosed buildingcan become a problem if any system leakage is


(1) Bi-fuel, where the vehicle can run eitheron natural gas or gasoline

(2) Dual-fuel, where the vehicle runs eitheron diesel only or diesel and natural gas,with the combustion of diesel used toignite the natural gas

(3) Dedicated, which runs entirely on natu-ral gas.

All three types can be manufactured from thestart to use natural gas by original equipmentmanufacturers (OEM), or converted from vehiclesoriginally manufactured to run on gasoline ordiesel only. Either way, there is an incrementalcost relative to vehicles using conventional liq-uid fuels. From the point of view of minimizingemissions, OEM vehicles (that is, vehicles thatare manufactured as NGVs at the factory level)are considered more suitable than convertedones, but their higher prices may make it diffi-cult to deploy them on a large scale in develop-ing countries. In 1998, 43 OEMs around the worldproduced NGVs, and 11 heavy-duty enginemanufacturers produced NG engines. Conver-sion of vehicles in poor condition, as well aspoor conversions, are two of the most seriouspotential problems in developing country cities,and could even defeat the purpose of switchingto NG. Impco Technologies, a major supplier offuel and electronic control systems for naturalgas to OEMs, estimates that 50 to 70 percent ofvehicles being converted in developing coun-tries will fail a good pre-conversion inspection(Impco Technologies 2000).

Because CNG vehicles are more expensiveto purchase than vehicles powered by liquid fu-els, for the NGV program to be financially sus-tainable in the long run, the incremental costmust be recovered in the form of lower operat-ing and maintenance costs. The lower cost inturn typically has to come from a much lowerprice of fuel per distance traveled. The viabilityof a NGV program therefore rests critically oninter-fuel pricing, and more specifically, the fueltaxation policy adopted by the government,

present. LNG poses an even greater safety threat.Appropriate roof venting is necessary to ensurethat natural gas exits the building. Garages forNGVs must be designed with good ventilationat the ceiling level.

The numbers of NGVs and refueling stationsare shown in Table 2. Over 1.5 million vehiclesrun on natural gas worldwide, fueling at morethan four thousand refueling stations. The larg-est NGV market is Argentina, followed by Italy,Pakistan and the United States. Close to half oftotal NGVs in the world are in Argentina. MostNGVs are light-duty vehicles converted fromgasoline. The number of NGVs in India has in-creased substantially since information was sup-plied in August 2000, on account of the SupremeCourt decision affecting Delhi. By mid-2001,there were close to 40,000 NGVs in Delhi alone(CPCB 2001). The United States has the largestnumber of refueling stations, with more thantwelve hundred.

When launching a NGV program, one logisti-cal problem is the balance between the numberof NGVs and refueling stations. Any imbalance—either in the form of over-supply of refuelingstations or a disproportionately greater numberof vehicles relative to refueling capacity—wouldresult in either very low returns for refuelingstation owners, tarnishing the image of the NGVindustry in the eyes of investors, or long queuesfor vehicle drivers, tarnishing the industry’s pub-lic image. Fleet operators with high usage ve-hicles may choose to set up their own refuelingstations. This is typically the case with transitbus operators in industrial countries that estab-lish filling stations at bus depots. In this case,economies of scale become an important con-sideration, since there is a minimum numberof vehicles that such a filling station shouldserve to be economic. For NG bus fleet op-erators, there are also economies of scale instaff training, fuel purchase, vehicle maintenanceand service (such as having a service contractfor the entire fleet).

There are three types of NGVs:


Table 2. International Natural Gas Vehicle Statistics

Country Vehicles Refueling stations Information as of

Argentina 668,480 923 May 01Italy 370,000 355 Mar 01Pakistan 200,000 200 Jun 01United States 102,430 1,250 Jan 01Brazil 80,000 131 Mar 01China 36,000 70 Jan 00Venezuela, Republica Bolivariana de 33,586 150 Jun 01Russian Federation 30,000 202 Sep 00Egypt, Arab Republic of 24,115 45 Jan 01Canada 20,505 222 Aug 00New Zealand 12,000 100 Aug 00Germany 10,000 146 Jan 01Colombia 10,000 28 May 01India 10,000 11 Aug 00Japan 8,053 138 Jul 01Bolivia 6,000 17 May 01France 4,550 105 Oct 00Trinidad and Tobago 4,000 12 May 01Malaysia 3,700 18 Oct 00Indonesia 3,000 12 Aug 00Australia 2,000 12 Nov 00Chile 2,000 7 May 01Sweden 1,500 25 Mar 00Bangladesh 1,000 5 Aug 00Great Britain 835 18 Aug 00Iran, Islamic Republic of 800 2 Aug 00Netherlands 574 27 Aug 00Spain 300 6 Aug 00Belgium 300 5 Aug 00Mexico 300 2 May 01Switzerland 270 14 Aug 00Korea, Republic of 245 3 Jul 01Turkey 189 3 Aug 00Thailand 184 1 Mar 01Austria 83 5 Aug 00Ireland 81 2 Sep 00Cuba 45 1 Feb 01Finland 34 5 Aug 00Czech Republic 30 11 Aug 00Nigeria 28 2 Aug 00Luxembourg 25 5 Aug 00Poland 20 4 Aug 00Norway 18 3 Aug 00Taiwan (China) 6 1 Nov 00Denmark 5 1 Aug 00Korea, Democratic People’s Republic of 4 1 Aug 00

Total 1,645,705 4,317

Source: International Association for Natural Gas Vehicles, http://www.iangv.org/html/ngv/stats.html#1


which determines the relative prices of fuels. InArgentina, for example, the retail price of CNGhas been historically about one-third of the priceof premium gasoline. As a result, car ownerswho switched from gasoline to natural gas real-ized 65 percent savings in fuel cost relative topremium gasoline from the beginning of thenatural gas vehicle program in 1985, rising to 70percent savings in 1999. In industrial countries,tax is a large fraction of the final price of liquidfuels, so it can be adjusted to favor one fuelover another. In 1999, tax on gasoline consti-tuted 67 percent, and that on diesel 59 percent,of consumer prices in the countries of theOrganisation for Economic Co-operation andDevelopment (OECD) (Bacon 2001). This is alsothe case in most developing countries with re-spect to gasoline, but much less so with diesel.As a result, the retail price of diesel has histori-cally been one-half that of gasoline or even lowerin countries such as Argentina, India, Indonesiaand Sri Lanka.

Natural gas has a much higher auto-ignitiontemperature than gasoline and diesel, making itsafer but also unsuitable for compression igni-tion, which is used in diesel-fueled vehicles. MostNG vehicles are conversions of existing liquidfuel vehicles. In the case of gasoline (spark ig-nition) engines, the conversions are generallybi-fuel. The conversion from diesel (compres-sion ignition) to NG is not straightforward. Thetwo main options are dedicated, involving con-version to spark-ignition, and dual-fuel, entail-ing co-existence of two fuel injection systemsand adding to the complexity of the engine.

REASONS FOR SWITCHING FUEL

TO NATURAL GAS

There are two principal reasons for switching tonatural gas. One is the significantly lower exhaustemissions, especially of particulate matter. Thisis the primary reason for the government’s pro-motion of natural gas buses in the United States,

as well as the Supreme Court decision in Delhi,India. If exhaust emission reduction is the pri-mary reason, then dual fuel transit buses do notachieve the objective all that well, because thestop-and-start nature of the urban bus drivingcycle means that the substitution of diesel bynatural gas is limited.

The second reason is diversification of en-ergy sources. This has in fact been the historicalreason for switching to natural gas. Worldwidenatural gas reserves are more abundant than oilreserves, giving greater potential to the use ofnatural gas. In 2000, the ratio of proven reservesto production of natural gas was estimated tobe 62 years, and that of oil 38 years (bp 2001).A country that imports oil, but has an abundantsupply of natural gas, may find it particularlyattractive to consider natural gas as a transportfuel in order to reduce its oil import bills.Bangladesh and Indonesia (where crude oil pro-duction will cease in less than a decade at thecurrent rate but abundant supplies of naturalgas remain) cite this as the reason for wantingto promote NGVs. There are also other ways ofusing natural gas in transport, such as conver-sion of natural gas to synthetic fuels (includingsuper-clean diesel, designated by the U.S. gov-ernment as an alternative fuel in 2000) or tomethanol or dimethyl ether. In New Zealandwhere the government aggressively promoted aNGV program after the oil price shock of theearly 1980s, the country’s own natural gas re-sources were used not only directly as a trans-port fuel but also as a feedstock for makingsynthetic fuels in the 1980s. The production ofsynthetic fuels from natural gas—which is notyet economic at the current world price of crudeoil and available technologies for convertingnatural gas to liquid fuels—has been commer-cially carried out and continues in South Africaand Malaysia today.

There are cases of oil-importing countrieswith indigenous natural gas reserves whereswitching to natural gas is not necessarilyfinancially favorable. Pakistan illustrates this


point. In Pakistan, as in the rest of South Asia,diesel has historically been priced at one-halfthat of gasoline. As a result, it is not uncommonto see light-duty gasoline vehicles converted todiesel. Partly as a result of this pricing policy,the consumption of gasoline is less than a fifthof that of diesel. While Pakistan imports the bulkof its diesel, it has recently become a net ex-porter of naphtha (which is used in gasoline

production). Because of Pakistan’s inter-fuelpricing, natural gas has displaced only gasolineand not diesel—it is much more attractive forvehicle owners to switch from gasoline to natu-ral gas than from diesel to natural gas—worsen-ing the supply-demand imbalance and forcingthe refineries (which are financially supportedby the government to some extent) to exporteven more naphtha at a loss.

15

Chapter 2

International Experience with NaturalGas Vehicles: Cases of Argentinaand New Zealand

T he worldwide population of natural gasvehicles grew at an annual rate of about5 percent between 1994 and 2000. Dur-

ing this period, the number of NGVs grew from250,000 to 620,000 in Argentina, the largest NGVmarket, corresponding to an annual growth rateof 16 percent. The growth in Italy, which hadthe same number of vehicles as Argentina in1994, was slower, increasing from 250,000 to320,000. In other countries, the NGV popula-tion declined, including New Zealand—once oneof the largest NGV markets in the world—andthe newly independent states of the former So-viet Union. This chapter reviews the factors af-fecting the growth or decline of NGV marketsthrough two case studies, Argentina (Francchia2000) and New Zealand (Harris 2000). While theNGV market in both of these countries has fo-cused exclusively on converting existing gaso-line vehicles to run on natural gas, generallessons can be learned to understand the natu-ral gas bus industry.

ARGENTINA

Background

Argentina is endowed with both oil and naturalgas. At the end of 2000, the ratio of proven re-serves to production was 10 years for oil and 20years for natural gas (bp 2001). The Liquid Fu-els Substitution Program was launched in 1984

to free up more oil for exports (oil is easier toexport than natural gas) and to increase fuel taxeson liquid fuels without provoking widespreadpublic protests by offering low price CNG as anautomotive fuel. By then, an extensive networkof natural gas pipelines reached most cities.

To start the program, two refueling stationswere established and a few government vehiclesand taxis were converted from gasoline to natu-ral gas. Because of the domestic economic situ-ation, no subsidies could be offered to “get theprogram off the ground.” The incentive forswitching to CNG was the very large price dif-ference between gasoline and CNG, ensuring65 percent savings in fuel cost by switching frompremium gasoline to CNG.

Safety, quality and other standards were de-veloped and enforced by the regulatory authori-ties for gas cylinders, conversion kits, conversionworkshops, compressors, dispensers, installationprocedures and so on. Internationally well-known certification agencies carried out the cer-tification.

In the late 1980s, the government began toincrease the retail price of diesel, aiming in thelong run to substitute diesel with natural gas.The primary objective of the Liquid Fuels Sub-stitution Program was in fact the substitution ofdiesel by natural gas in public transport vehicles.The fiscal policy changes needed to achieve thissubstitution, however, were not implemented. Theretail prices from recent years are shown in Table3. Diesel has historically been and continues to


be comparable to CNG in price. After takingengine efficiency into account, the price of die-sel was lower than that of CNG in May 1994,comparable to that of CNG in December 1998,and rose above that of CNG only in December1999, so that diesel has actively competed withCNG in the light-duty vehicle market to replacegasoline. Half of new taxis in Argentina arediesel-powered.

The evolution of the market shares of differ-ent fuels in the 1990s is given in Table 4. Themarket share of gasoline steadily declined, thatof CNG increased until 1997, and that of dieselfluctuated until 1997 after which it saw a markedincrease at the expense primarily of gasoline,but also of CNG.

Between 1985 and 1999, direct investment inthe NGV market totaled US$1.5 billion. The con-verted vehicles operate as bi-fuel vehicles. Inthe early years of the NGV program, it took lessthan two years to pay back the vehicle conver-sion. The number of months it takes to pay backthe conversion cost of US$1,200 as a function ofannual kilometers (km) traveled for the premiumgasoline and CNG prices effective December1999 is shown in Figure 1. Today, the NGV in-dustry generates US$0.65 billion worth of busi-ness per year. About 1.5 billion m³ of naturalgas is sold annually as a transport fuel, approxi-mately equal to the amount of gas consumed inthree 500-megawatt power plants. The numberof vehicles converted has stabilized at about5,000 a month.

In contrast to gasoline vehicles, the econom-ics for converting diesel vehicles to NG are muchless favorable even for intensively driven ve-hicles. Take, for example, a bus driving 120,000

Table 3. Representative Fuel Prices

Premium gasoline Diesel Natural gas(US$/liter) (US$/liter) (US$/m³)

Date Total Tax component Total Tax component Total Tax component

May 1994 0.751 0.369 0.269 0.033 0.256 0.095December 1998 0.906 0.588 0.403 0.180 0.311 0.106December 1999 1.044 0.608 0.499 0.211 0.331 0.082

Source: Francchia 2000.

Table 4. Percent Market Share of Liquidand Gaseous Fuels

Year Gasoline Diesel CNG

1990 42 56 21993 40 55 51995 37 57 61997 35 58 71999 29 65 6

Source: Francchia 2000.

km a year. Assuming a vehicle purchase pricedifference of US$22,000 and fuel prices as ofDecember 1999, it takes 59 months to recoverthe incremental vehicle cost based on fuel costsavings, or close to five years. The fuel prices asof December 1998 would have increased thepayback period to 29 years, considerably be-yond the useful life of the vehicle.

One of the difficulties in launching a NGVprogram is balancing the numbers of NGVs andrefueling stations. Inadequate refueling infra-structure was partly responsible for conversionback to gasoline from CNG in Bangladesh inthe 1990s, and long queues at CNG refuelingstations in Delhi are a significant source of dis-satisfaction among vehicle drivers today. InArgentina, the NGV market has been developedexclusively by the private sector, with many play-ers entering the NGV market—refueling; manu-facture of compressors, dispensers and gascylinders; and manufacture and installation ofconversion kits. In setting up the refueling in-frastructure, those who had not been previouslyinvolved in the fuel retail business opened CNGrefueling stations, as did oil companies. The pay-back period for an independent operator of arefueling station was approximately three years

17International Experience with Natural Gas Vehicles

in the 1990s. By establishing dual fuel stations(selling both liquid fuels and CNG), oil compa-nies went around the regulation prohibitingrefueling stations to be set up within 2 km ofeach other.

Lessons from Argentina

The NGV program in Argentina is the most suc-cessful in the world, measured in terms of NGVpopulation. This program has focused almostexclusively on converting existing gasoline ve-hicles to CNG, taking advantage of the large pricedifference between the two fuels. The large pricedifference in turn is provided by the fuel taxstructure. Aside from this indirect support, thegovernment has given no subsidies in the formof financial incentives to the CNG industry, mak-ing the CNG program viable in the long run.The CNG industry in Argentina today is export-ing ISO 9000 certified compression and dispens-ing equipment, gas cylinders and conversion kitsto other countries in Latin America and Asia. Atthe same time, the program has made no dentin the automotive diesel market, which is be-ginning to threaten the CNG market. Heavy-dutypublic transport vehicles remain entirely diesel

fueled. A review of inter-fuel taxation policy aswell as vehicle tax policy would be needed ifthe growth of the automotive diesel market is tobe halted in the coming years.

NEW ZEALAND

Background

New Zealand was a world leader in CNG ve-hicles in the middle of the 1980s. The sale ofnatural gas as CNG peaked in 1985 at close to150 million m³ a year. Between 1979 and 1985,the number of NGVs doubled every year. By1986, CNG represented one-tenth of the fuel usedby spark ignition engine vehicles (that is, gaso-line engine vehicles) in the North Island—theonly island where natural gas is available.

The oil crisis of the 1970s affected NewZealand, which was importing nearly all of itstransport fuels, prompting the government toseek alternative forms of energy. A major off-shore gas field had been discovered in 1969.While the field was developed for power gen-eration, it became apparent by the late 1970sthat demand for power was well below the fore-

0

50

40

30

20

10

0

20,000 40,000 60,000 80,000 100,000

Payback in months

Kilometers traveled per year

Figure 1. Payback for Conversion from Fuel Cost Savings in Months

Source: Francchia 2000 and author's calculations.


casted levels, requiring much less gas than origi-nally envisaged even though the governmenthad already signed a take-or-pay agreement. Thegovernment therefore evaluated CNG as an al-ternative use of natural gas, and concluded thatsubstitution of gasoline by CNG would helpaddress two economic problems: balance ofpayments and unemployment. Urban air pollu-tion was not a consideration in the government’sdecision to launch a NGV program at the time.

The government sponsored extensive inves-tigations into the impact of adopting CNG as atransport fuel starting in 1974. Based on the find-ings and recommendations, the government in1979 set a target of 150,000 CNG vehicles by1985, subsequently revised to 200,000 by 1990.Additional recommendations for the use of natu-ral gas included the construction of a syntheticgasoline plant and a chemical methanol plant,both using natural gas as the feedstock. At thetime, there were about 100 gas utility vehiclesusing CNG and only two refueling stations inNew Zealand. Few people had technical exper-tise. The Ministry of Energy was formed in 1978,and it was not in a strong administrative posi-tion to coordinate the implementation of theCNG program in 1979. In response, the govern-ment established the CNG Coordination Com-mittee (CCC) to coordinate efforts within thegovernment and the private sector. The Ministryof Energy became the lead agency in 1981, by

which time both the CNG program and the min-istry were well established.

The government also began to offer financialincentives for vehicle conversion and refuelingstations in the form of NZ$200 grants for con-version kits, 25 percent grants for mechanicalequipment in refueling stations and loans forrefueling stations. The government mountedprograms for the implementation of the NGVprogram: providing training, establishing stan-dards for vehicle conversion and refueling sta-tions, and mounting public awarenesscampaigns. A major boost to the CNG industrywas the decision of the government to requireits own fleet to convert to CNG. Within a year,refueling stations became grossly overloaded,and half-hour queues were common. A list ofsteps taken by the government is summarizedin Table 5.

By the middle of 1980, it became clear that aprice differential of 50 percent between gaso-line and CNG was not sufficient to achieve theconversion target set by the government. At thesame time, various technical problems arose, insome cases giving rise to extremely adversepublicity. The rate of conversion fluctuated er-ratically. A rapid market survey conducted atthe end of 1980 convinced the government thatfurther incentives were needed. The package ofincentives, announced at the end of 1980, in-cluded accelerated depreciation for vehicle

Table 5. Steps Taken by the Government of New Zealand

Year Action

1978 Funding of research and development for evaluating CNG.1979 Formulation and acceptance of the implementation plan. Setting of targets and

incentives. Establishment of CNG Coordination Committee.1979-1985 Establishment of regulations and standards, and infrastructure for inspection. Various

promotional and marketing activities. Conversion of government vehicle fleet to CNG.1979 onwards Training programs for installers.1980 Market survey. Modification of incentives.1980 onwards Funding of engine and related research.1981 Market survey.1982 Modification of incentives.1983 100 percent government loans for vehicle conversions.1984 Market survey. Modification of incentives. Election of new Labor Government.1985 Target and incentives abandoned by the new Labor Government.

Source: Harris 2000.


conversions (then costing about NZ$1,000 pervehicle) and changed rules for the 25 percentrefueling station grant. There was an immedi-ate response, with the rate of conversion nearlydoubling between the latter half of 1980 andthe first half of 1981. Additional surveys wereconducted in 1981 and 1984 to determine theappropriate level of financial incentivesneeded.

The above incentive schemes were followedin 1982 by the introduction of industry-fundedCNG vouchers entitling voucher holders toNZ$300 of free CNG, and in 1983 a 100 percentgovernment loan for vehicle conversions tookthe place of the accelerated depreciation pro-gram described above. Between July 1984 andOctober 1985, the price of CNG was about 40 to50 percent that of gasoline on an energy equiva-lent basis. During that period, the rate of con-version rose sharply, fell, and rose sharply again,challenging the ability of the industry to under-take high quality conversions. By 1985, over100,000 CNG conversion kits had been sold,nearly all converting gasoline to CNG. Few die-sel vehicles were converted. The gross incomeof the CNG industry in 1984 was approximatelyNZ$84 million (US$49 million) and net foreignexchange savings amounted to NZ$30 million(US$17 million). Between 1979 and 1985, thenet cost to the government of the various incen-tive schemes and loans was in excess of NZ$20million (or about NZ$200 per vehicle), in addi-tion to administrative costs such as research anddevelopment, promotion and servicing of com-mittees.

A number of implementation issues arose andwere handled with varying degree of success.

Technical. A wide range of technical issuesrelated to vehicle conversion and refueling sta-tions were investigated and handled. Some wereknown from the beginning and required inves-tigation followed by a decision, such as settingthe maximum cylinder filling pressure. Othersarose as experience with CNG vehicles pro-

gressed, and some were unforeseen (for ex-ample, the maximum cylinder filling pressureset in 1979 proved to be too low and was laterraised). Much of the original technology wasimported, especially from Italy and the UnitedStates. The constraints faced by New Zealandincluded the fact that there was (and is) no ve-hicle manufacturing capacity but only vehicleassembly, nor does New Zealand possess anindustry to manufacture major items of CNGequipment.

Institutional. The CNG program was at firstgreeted with skepticism, if not outright opposi-tion. Against this setting, the chair and execu-tive officer of the CCC effectively became theproduct champions. The CCC had an influenceon almost all the major government decisionsrelated to CNG except the financial incentives.The role of the CCC was initially to lead andcajole various agencies, and after greater accep-tance of CNG and cooperation, to promote andmarket the product. The CCC had no formal le-gal or administrative status. It relied on persua-sion at the beginning. In time, the Ministry ofEnergy and the industry gave weight to the sug-gestions made by the CCC.

Setting standards for vehicle conversion andrefueling stations became a key activity withinthe Standards Association of New Zealand. Ac-ceptance of gas cylinders manufactured in Italyby the Dangerous Goods Inspector of the De-partment of Labor required extensive translationand interpretation of the Italian cylinder designrules. Establishing training courses for mechan-ics needed action by the Motor Industry Train-ing Board and the Department of Education. TheNew Zealand Energy Research and DevelopmentCommittee, a government agency, provided keyinputs from the outset. It funded the originaltechnology assessment in 1978 and preparedthe implementation plan in 1979. It funded sev-eral research projects, especially those directedat how CNG vehicle engines in New Zealandperformed.


A small number of government personnel in-volved with CNG in the beginning stayed withimplementation and provided continuity, greatlyassisting with progress of the program. Theirprincipal interest was to ensure that New Zealandhad a viable alternative to gasoline as a trans-port fuel. In addition, a small number of peoplefrom private sector firms had a similar determi-nation to make CNG a success because of thepotential profits the CNG business could bringto their firms. The success of the CNG programcan be said to be due in large part to the effortsof these key government and private sectorpeople.

Economic. It goes without saying that, givenenough incentives, vehicle owners will switchto CNG. The government of New Zealand intro-duced a number of financial incentives that per-suaded vehicle owners and businesses to convertvehicles and construct refueling stations. Theprice difference between gasoline and CNG wasadequate to give a payback period of 18 monthson the investment for vehicle owners spendingin excess of NZ$35 per week on gasoline at thetime. The industry-funded CNG voucher schemeprovided an additional incentive. The capital costof conversion was covered by a 100 percentgovernment loan. Government grants and loanshelped establish refueling stations.

The market survey conducted jointly by thegovernment and the private sector found thatthe quality of vehicle conversions had not beengood, pointing to a need for quality assuranceand warranty for vehicle owners. In fact, CNGwas seen as a second-rate fuel used only be-cause it cost less than gasoline. Vehicle ownersconsidering conversion weighed the problemsand disadvantages of CNG vehicles against themuch lower fuel price. Refueling was not seenas a major problem because 450 refueling sta-tions had been established by 1984.

The new Labor Government elected in 1984adopted an economic policy of deregulation andliberalization, withdrawing the incentives offered

for conversion and refueling stations. From 1985,conversions rapidly declined to almost zero andthe consumption of CNG fell gradually as exist-ing CNG vehicles ended their normal useful life,as did the number of refueling stations. A littleover 10,000 CNG vehicles remained in the na-tional fleet in the year 2000, a significant de-cline from the peak of 110,000.

Lessons from New Zealand

The CNG program in New Zealand developedagainst the backdrop of very high internationaloil prices following the Iranian revolution andan indigenous supply of natural gas with de-mand not matching the amount that the govern-ment had agreed to “take or pay.” In response,the government of New Zealand took the leadin promoting the CNG vehicle program aggres-sively. It sponsored research, prepared the imple-mentation plan, and coordinated the entireprogram. Most important, it provided generousfinancial incentives, so that the number of CNGvehicles doubled every year, seriously stretch-ing the ability of the industry to cope. The in-dustry was so preoccupied with meeting thedemand for conversion that quality at times be-came a secondary priority, resulting in a poorperception of CNG vehicles in some quarters.

When the new Labor Government began toderegulate the economy, withdrawing supportfor the CNG industry in the form of financialincentives, the market essentially died. A CNGprogram that relies heavily on government sub-sidies, as in New Zealand, is not likely to besustainable in the long run. Inter-fuel pricing inNew Zealand today suggests that the world oilprice must rise above US$30 per barrel beforeCNG becomes commercially viable without gov-ernment support. The New Zealand experiencesuggests that the price of CNG should be nomore than half of the retail price of gasoline it issubstituting. Further, if the CNG price is 30 per-cent of the gasoline price, no direct support isnecessary, but at 50 percent some government


support in the form of financial incentives isbelieved to be needed.

OTHER INTERNATIONAL EXPERIENCES

Before the 1990s, many countries turned to natu-ral gas as a means of achieving greater energyself-sufficiency, security of supply or lower fuelimport bills. Environmental advantages of NGVs,especially relative to diesel, began to play aprominent role in the 1990s. When gasoline ve-hicles were converted to natural gas, an unpleas-ant surprise in industrial countries was thatemission tests showed that converted vehicleswere more polluting than recent model-yeargasoline counterparts. What emerged is thatnewer vehicles equipped with electronic emis-sions control packages were less amenable to asuccessful aftermarket conversion. In fact, themost advanced commercially available gasolineengine vehicles today are extremely clean andNGVs have no obvious emissions advantage overthem. However, NGVs do produce lower green-house gas emissions over the vehicle’s life cycle,which is an increasingly important consider-ation given the rising contribution of the trans-port sector to overall greenhouse gas emissionsin most countries. In contrast, with older ve-hicles, or in markets where manufacturers arenot required to provide sophisticated elec-tronic and emissions controls, or where leadedgasoline is still extensively used, the conver-sion to natural gas can provide immediate emis-sions improvement. One clear-cut example isthe conversion of gasoline vehicles running onleaded gasoline to CNG in Pakistan: CNG con-tains no lead, so switching from leaded gasolineto CNG eliminates lead emissions. As for diesel,particulate emissions from NGVs are much lowerthan those of conventional diesel, often by afactor of ten or more. Only the so-called cleandiesel with sophisticated after-exhaust treatmenttechnology and ultra-low sulfur diesel can be-gin to match the emission levels of their naturalgas equivalents.

In Chile, where tests for converting to CNGhave been carried out with taxi fleets, the con-versions failed because of a poor choice of af-termarket gasoline-to-CNG conversions; theoption of a fully factory-built CNG car wouldhave been more satisfactory. It is very impor-tant, when retrofitting existing vehicles, to carryout conversion properly, ensuring customer sat-isfaction as well as achieving the expected emis-sions reductions.

The greatest barrier to the expansion of theNGV market has been the high cost of refuelingstations, vehicle conversion, and OEM vehicles.In North America and the EU, successful con-versions have been difficult and costly becauseOEM vehicles are generally recognized to benecessary to ensure minimal emissions. As aresult, NGVs have been confined mostly to high-usage vehicles.

ROLE OF GOVERNMENT

The cases of Argentina and New Zealand high-light a number of issues related to the role of agovernment in launching and sustaining a NGVprogram, and that of government support inparticular. The government can assist in a num-ber of ways.

Potential Government Assistance

� Establishing a proper regulatory framework isone of the principal roles of the government.The government should ensure that there is alevel playing field, players are encouraged toincrease efficiency and quality of service andproducts through competition, and monitor-ing and enforcement of regulations and stan-dards are adequate.

� Establishing safety and performance stan-dards is another important government func-tion. Both Argentina and New Zealand movedquickly to address this aspect, although themaintenance of performance standards wasless than satisfactory in the latter.


� Adopting an inter-fuel taxation policy favor-able to automotive natural gas is necessary ifthe NGV program is to be viable and sustain-able on a commercial basis. In Argentina,gasoline is so heavily taxed that CNG is com-mercially competitive. In neither country hasdiesel been taxed to the extent necessary topromote conversion from diesel to CNG. InNew Zealand, one of the key incentives forconversion came from government subsidies,so that when the subsidies were withdrawn,inter-fuel price differences alone could notsustain the CNG program. A further questionis the extent to which the retail prices shoulddiffer. For a given payback period, high-us-age vehicles need a smaller price differencethan lower-mileage vehicles. If all vehiclesincluding private passenger cars are targetedfor conversion, the price difference requiredwould be much greater than if only high-us-age commercial vehicles such as taxis anddelivery vans are targeted.

� Providing subsidies was the policy aggressivelypursued by the government of New Zealand.In the early days of a NGV program, the in-fant industry argument may justify subsidies.For example, to break the logjam in whichcar owners wait for adequate refueling infra-structure before investing in a fuel switch,while business enterprises wait for a sufficientnumber of converted vehicles before invest-ing in refueling stations, the government mayconsider subsidizing startup costs. However,as the case of New Zealand demonstrates, largesubsidies are unlikely to be sustainable in thelong run, threatening the survival of the NGVprogram. In the words of Robert Cummingwho spoke on behalf of the International As-sociation for Natural Gas Vehicles in MexicoCity in 1997, “Governments that believe thatall they need is a two- to three-year kickstartare wasting their time and money” (Cumming1997). It would be preferable to provide mod-est but consistent support over a long periodof time than large subsidies that are reducedsignificantly or withdrawn altogether after a

few years. Heavy subsidies may also lead toserious market distortions, such as over-sup-ply of refueling stations.

� Providing non-monetary incentives is anotheroption. Examples include reduced frequencyof required emissions inspection tests or theright to drive a CNG vehicle in high-occu-pancy vehicle lanes or on days when othervehicles are not permitted (as in cities thatban vehicle usage on certain days to reduceair pollution). Such incentives alone wouldnot induce vehicle owners to switch to natu-ral gas, but coupled with other incentives(most importantly fuel cost savings), theycould play a useful role.

� Mandating conversion to natural gas is not astep that should be taken lightly, especially iffinancial and logistical (fueling and drivingrange) burdens are anticipated to be great onvehicle owners. The Supreme Court decisionimposed in Delhi for buses is one example.The New Zealand government’s decision toconvert government fleets to CNG in onesense falls under this category. An indirectway of mandating conversion is to set emis-sion standards that can be met only by NGVs.In the United States, the South Coast Air Qual-ity Management District (SCAQMD) in Cali-fornia has recently banned diesel buses infavor of NG and other alternative fuel engines.

� Acting as a champion is a consideration forthe government, especially in the early daysof the NGV program. The government canpublicize the benefits of NGVs, perhaps us-ing prominent senior officials to reinforce themessage. It is equally important for the pri-vate sector to assume this role.

Designing Fuel Tax

The issue of inter-fuel taxation is a complex oneand is beyond the scope of this report. How-ever, a few general principles from tax theorymay be outlined here. To devise an optimal taxscheme, which would enable the governmentto raise sufficient revenues while minimizing the


loss of consumers’ welfare from the higher pricesthey would have to pay because of the taxes,the following rule is often taken as the startingpoint: if a certain set of conditions are met1 ,then no intermediate goods should be taxed,and the tax rates on final consumption goodsshould be inversely proportional to their ownprice elasticities of demand. Thus, if consumersare likely to cut back consumption markedly inresponse to a price increase (as in the case ofcertain luxury goods), that item should not betaxed much, but if consumers are likely to con-tinue to consume only slightly less on accountof the price increase (as in the case of such staplefood items as rice or maize), then the item shouldbe taxed relatively more. Under these conditions,because diesel used in freight and passengertransport, industry, and agriculture is an inter-mediate good, diesel for these purposes shouldnot be taxed.

However, the above conditions are not satis-fied:

� Vehicles in cities cause congestion, dieselemissions are harmful to public health, andall vehicles, but heavy-duty vehicles consum-ing diesel in particular, damage roads, so thatthere is an external cost associated with theuse of diesel (for productivity loss due to con-gestion, additional healthcare costs and ex-penditures for road maintenance)

� Diesel and gasoline are substitutes for light-duty vehicles in the long run so that taxingdiesel little and gasoline much more wouldresult in an automotive fuel switch out of gaso-line to diesel

� In many developing countries not all finalgoods can be taxed, so that taxes on petro-leum products, which are relatively easy tocollect, become an important source of gov-ernment finance, especially in low incomecountries

� A number of markets in developing countrieshave distortions that impede perfect or evennear-perfect competition.

All these trends argue for taxing diesel evenwhen it is an intermediate good. That gasolineand diesel are substitutes in the light-duty ve-hicle category is a particularly strong argumentfor making their tax levels comparable, or elsethe fuel that is taxed less (almost universally die-sel) will be consumed more, eroding the taxbase and requiring higher tax rates elsewhereto collect the same amount of money.

Yet another consideration in designing tax isequity—items that the poor consume dispropor-tionately more than the rich as a share of theirtotal expenditures (such as food) should be taxedless than the above “inverse elasticity” rule wouldimply so as to lessen the tax burden on the poor.Conversely, for goods consumed more by therich than the poor, such as gasoline, the tax rateshould be higher. Where the impact of an in-crease in the price of diesel on household ex-penditures has been studied, the effects havebeen found to be regressive—that is to say, thetotal expenditures of poor households rise morein percentage terms than those of the rich whenthe price of diesel is raised—although the mag-nitude of the impact is not large, remaining ofthe order of a couple of percentage points. Thiswould argue somewhat for not raising the taxon diesel as much as the above factors mightsuggest. This is one reason why some govern-ments view diesel as a “social” fuel, limiting taxon diesel compared to gasoline, which is seenas a fuel for the rich, since only better-off fami-lies can afford to purchase motorized vehicles.Nevertheless, the equity argument alone wouldnot justify keeping the end-user price of dieselat half that of gasoline as seen in a number ofcountries.

Natural gas used in the transport sector is nodifferent than liquid fuels from the point of viewof tax theory with one exception: the environ-mental externality is lower relative to old tech-nology gasoline vehicles, and considerably lowerwith respect to conventional diesel. To set thetax level capturing externalities would require aknowledge of contributions of vehicles with


different fuels to the overall air and noise pollu-tion, and health and other damages associatedwith each fuel. This level of information is sel-dom, if ever, available in most developing coun-try cities.

It is important to note that the incrementaltax adjustment to deal with the externality shouldbe applied to that good only—there is no rea-son to tax complements more heavily or sub-stitutes less heavily independent of their ownpolluting characteristics. In the case of fueltaxes, this means that government should taxdiesel more, and not lower the rate of tax onnatural gas. Further, additional considerations in-clude other externalities associated with NGVs—congestion and damage to roads (which wouldincrease because NGVs are heavier)—as well asthe fact that natural gas and liquid fuels are closesubstitutes. Subsidized natural gas is made avail-able in a number of countries. A prime exampleis natural gas sold to the fertilizer industry. It isimportant to have market-based natural gas pric-ing rather than government-determined below-market pricing for the long term viability of theNGV program. Subsidizing natural gas (that isto say, selling it below cost) to promote its usein the transport sector, a position promoted bysome, ignores these widely accepted principlesof optimal tax theory.

Another important point is that fuel taxationis a poor proxy for an efficient externality charge,because air pollution or congestion is a highlylocalized phenomenon, while fuel tax is usuallyset at a national level. From the point of view oftaxing environmental health damages, emissionsin densely populated areas need to be taxed,which a tax on diesel does not capture well.The option of heavily taxing urban buses fueledby conventional diesel, which may be more tar-geted, invites other problems: buses are oftenthe transport mode used by the poor, while aheavy tax on diesel buses would result in higherbus fares; and such a vehicle tax scheme mayalso eliminate transit buses altogether in favorof numerous mini-buses, which may be more

difficult to regulate and control for emissions.These issues point to the complexity of settingfuel and vehicle taxes in such a way to mini-mize distortions and maximize welfare.

Another reason cited for consideration in set-ting inter-fuel taxation is balance of paymentsfor countries that have indigenous sources ofnatural gas and that import crude or refined prod-ucts. However, if the exchange rate is fully de-termined by market forces so that it reflectsopportunity costs, there is no reason to differ-entiate taxes to save imports. A related issue isdiversification of energy sources so as to mini-mize the impact of possible future price hikes.This may justify differentiated taxation to a de-gree, although not so much as to give incen-tives to switch entirely from liquid fuels to NG.

Given that diesel is taxed little or even subsi-dized in many developing countries, conversionfrom diesel to natural gas would become eco-nomic only if diesel itself or diesel vehicles aretaxed much more. While there may be a num-ber of good reasons why the retail price of die-sel relative to gasoline should be raised in thelong run, there would nevertheless be a signifi-cant impact on other uses of diesel—in rail trans-port, agriculture, and industry, for example. Oneway of addressing this is to give rebates on thediesel tax to industrial and agricultural users ofdiesel. In any event, promotion of NGVs is un-likely to play a dominant role in determiningdiesel taxation. In practice, a combination of anumber of instruments are likely to be neededto achieve multiple objectives, including taxingitems that cause negative externalities; moreuniform taxes across different fuels that are sub-stitutes; tax rebates to industrial users of fuels;higher taxes on diesel vehicles, particularly thoseused primarily in intracity transport; and targetedsubsidies for the poor to compensate for higherexpenditures resulting from increased taxes, tomention a few.

Because gasoline is already taxed much morein most developing countries, if CNG has anychance of success on a commercial basis, it is as


a gasoline substitute. A large price differencebetween gasoline and CNG is currently achievedby taxing CNG much less. Several issues needto be considered in this case:

� If gasoline is effectively the sole source of taxrevenue from refined products (because allother fuels are taxed little or subsidized), thegovernment may not welcome a successfulCNG program whereby consumers shift fromrelatively heavily taxed gasoline to essentiallyuntaxed CNG. This would be particularly aconcern in low-income countries where taxrevenue from hydrocarbons accounts for asignificant fraction of the government’s totaltax take.

� If the CNG program is so successful that asizable portion of the gasoline market is re-placed by CNG while the automotive dieselmarket is untouched, the resulting productslate (with a very low gasoline-to-diesel ra-tio) will be difficult to manage for countrieswith refineries.

� If taxing CNG little does not provide suffi-cient financial incentives for conversion, thegovernment might consider increasing the taxon gasoline further (which in turn could fur-ther reduce the gasoline-to-diesel ratio fordemand), or reducing the tax on CNG, or both.

If the tax difference between CNG and gaso-line is to be widened at all, it would probablymake sense to target high-usage vehicles only.Leakage—diversion of CNG to users not tar-geted by the government—is unlikely to be aserious concern for two reasons. First, unlikeliquid fuels, natural gas is much more diffi-cult to transport, so that diversion to non-au-tomotive users from refueling stations wouldnot be simple. Second, since the tax schemewould target the price difference betweenCNG and gasoline to be at the level that wouldmake conversion to CNG financially attrac-tive only for high-usage vehicle owners, lowermileage vehicle owners would not benefitfrom converting to CNG to take advantage ofthe price difference.

NOTE

1. The conditions are that the economy is per-

fectly competitive, so that tax changes will be passed

on fully to consumers; there are no externalities as-

sociated with the consumed goods (such as air pol-

lution or congestion); all consumers are identical; all

final consumption goods can be taxed; and the items

taxed at different rates are not substitutable (such as

butter and margarine).

27

Chapter 3

Comparison of Natural Gasand Diesel Buses

T he conventional diesel engine is veryenergy efficient and reliable. One of thegoals of heavy-duty NGVs is to have

diesel-like efficiency and reliability. There areconsistent reports that the performance of thegeneration of NG buses from the early 1990swas less than satisfactory. The latest modelheavy-duty NGVs are much improved, but theengine technology still needs some refinements.

Comparison of heavy-duty NG and diesel ve-hicles is made difficult by the fact that both areevolving technologies. In the area of energy ef-ficiency and reliability, conventional diesel ve-hicles have reached a mature stage and themajority of the current development efforts fo-cus on reducing exhaust emissions. NGVs startwith an inherently lower emissions base, anddevelopment efforts have focused just as muchon improving its fuel economy and reliability.

In comparing NG and conventional fuel ve-hicles, it is important to clarify what is beingcompared. Comparing the state-of-the-art NGVswith conventional diesels, or vice versa, couldfavor one over the other. While comparison ofthe latest NG and diesel technologies may be ofprimary interest in North America, EU and otherindustrial regions of the world, this comparisonmay not be relevant to policy discussions indeveloping countries. Unfortunately, nearly allpublished data are from industrial countries,making it difficult to draw conclusions for de-veloping countries. Another important consid-

eration for developing countries is the emissionsand performance characteristics of vehicles asthey age, especially in countries that have poorcultural acceptance of regular maintenance. Thisinformation is not widely available for NGVs, asmany published studies have measured emis-sions from relatively new vehicles.

PERFORMANCE

There are two types of engines, compressionignition and spark ignition. All diesel enginevehicles have compression ignition, while gaso-line and dedicated NGVs have spark ignition.Compression ignition engines rely on self-igni-tion upon injection into hot, high pressure com-pressed air, and enjoy a number of advantagesover spark ignition engines, which ignite a ho-mogeneous and compressed pre-mixed mixtureof fuel and air with a spark. These advantagesinclude lower fuel consumption, longer life andsafer operation. Compression ignition enginesrun “lean,” or at a high air-to-fuel ratio, so thatcombustion occurs in the presence of excessair. In contrast, spark ignition engines typicallyrun “stoichiometric,” meaning that the air-to-fuelratio is adjusted so that the amount of oxygen inthe air is exactly that needed to combust all hy-drocarbons in gasoline. Vehicles with three-waycatalytic converters must use a stoichiometricmixture. The compression ratios of diesel engines


at 15-to-1 to 20-to-1 are much higher than thoseof gasoline engines at 8-to-1 to 10-to-1. Com-pression ignition engines do not have throttle,thereby reducing pumping losses. The lean-burncharacteristics, the lack of a throttle and a highengine compression ratio help to increase theefficiency of a diesel engine, leading to supe-rior fuel economy. Because dedicated NGVs haveto use spark ignition, all these advantages ofdiesel engines have to be sacrificed to someextent, although the high octane number of natu-ral gas compensates for some of the engine com-pression ratio loss in going from compressionto spark ignition in the case of OEM vehicles.Dual-fuel NGVs use a compression ignition.

Compression ignition engine components aretypically more robust than corresponding sparkignition engine components, increasing enginelife before overhaul or replacement significantly.The lean burn characteristics of diesel enginesprovide cooler exhaust temperatures, helping todecrease engine wear. Because compressionignition engines do not have ignition systems,there are no spark plugs or other ignition sys-tem components to clean and replace. However,the high pressure fuel injection system does re-quire maintenance.

There are two options for the air-to-fuel mix-ture ratio of heavy-duty NG vehicles: lean orstoichiometric. Cummins—the world’s largestdesigner and manufacturer of diesel engineswhich also offers the most widely fitted propri-etary CNG engines in Europe—is committed tolean burn NG engine technology for fueleconomy reasons. Others, such as Fiat-Iveco,have been developing stoichiometric NG units.In industrial countries, heavy-duty NGVs oper-ate either with stoichiometric mixtures and three-way catalysts, or with lean mixtures andoxidation catalysts. Stoichiometric units enablereducing emissions to extremely low levels andgive better drivability. The benefits of operatingin the lean-burn mode include greater enginedurability, higher fuel economy and higherpower output if turbocharging is used. Lean-burn

operation, however, has higher NOx and meth-ane emissions, is more sensitive to gas compo-sition variations, and can also lead to fastererosion of the spark-plug electrodes, needingreplacement as frequently as 300 hours (Nylundand Lawson 2000) as opposed to a durability of48,000 km required in the United States for gaso-line engines.

The maximum efficiency of a NG engine issome 10-15 percent lower than that of a gooddiesel engine. In practice, the difference maybe even greater. The energy consumption of aheavy-duty vehicle is estimated to increase, inmost applications, by 20-35 percent after con-version from diesel to natural gas. (Nylund andLawson 2000). Old conversion technologies de-crease fuel economy by as much as 25-40 per-cent. The fall in fuel economy arises from lowercompression ratios and throttling losses of NGVengines, and the additional weight of gas cylin-ders for fuel storage. The thermal efficiency is39 percent for a typical lean burn NG engine.As a comparison, new truck and bus diesel en-gines in the EU achieve thermal efficiencies ofaround 46 percent.

The driving range of NG buses is smaller thanthat of diesel buses. There is a trade-off betweencylinder weight and the desired range: the lowerthe cylinder weight for a given material of con-struction, the higher the fuel economy, but alsothe lower the range. The driving range of CNGbuses is often less than two-thirds of diesel buses.In the United States, driving ranges are of theorder of 400 km for CNG and over 650 km fordiesel buses. This has presented problems tosome bus operators who have had to arrangemid-day fueling (Montgomery County Transit,Maryland) and take other steps to ensure thatbuses do not run out of fuel. (Greater ClevelandRegional Transit Authority in Ohio reports thatNG buses have suffered out-of-fuel problems.)Similarly, TransAdelaide in Adelaide, South Aus-tralia, found that the range of NG buses was 11hours against their daily shifts of 15 hours, re-quiring the organization of mid-shift refueling.

29Comparison of Natural Gas and Diesel Buses

Diesel buses could complete the shift withoutrefueling. In contrast, buses for the Bangkok MassTransit Organization in Thailand travel about 220km a day on two routes, so that the driving rangeis not a critical issue.

EMISSIONS

The most visible and familiar emission from die-sel engines is the smoke trail produced whenthe vehicle operates under load. Consisting ofsolid particles and liquid droplets, smoke fromdiesel engines can be blue-white or gray-blackin color. Blue smoke is typically caused by thecombustion of lubricant found in the combus-tion chamber due to poor piston ring sealing orvalve guide wear. NGVs are also a source ofparticulate emissions, although at a much lowerlevel. White smoke is generated when the com-bustion temperature during fuel injection is toolow, which happens typically during transientoperation when starting, especially during coldweather or at high altitudes. White smoke canalso be produced when the injection timing isoverly retarded or when the compression ratiois too low. Gray-black smoke, consisting of car-bon particles, is generated when the engine isoperating at or near full load and too much fuelis injected or when the air intake is partiallyobstructed (for example, if the air filter is dirty).Gray-black smoke results from poor maintenanceof air filters and fuel injectors, or from improperadjustment of the fuel injection pump. Exces-sive smoke from diesel engines usually suggestsa loss in thermal efficiency, power output andfuel economy (Maxwell and Jones 1995).

Dedicated NGVs can enjoy a considerableexhaust emissions advantage over conventionaldiesel engine vehicles. In particular, visiblesmoke is virtually eliminated. A comparison ofemission test results of comparable diesel anddedicated CNG buses is shown in Table 6. Thereductions in emissions in going from conven-tional diesel to dedicated OEM CNG buses are

dramatic. The emission characteristics of dual-fuel vehicles depend in part on the extent towhich diesel is substituted by natural gas overthe engine operation range. As mentioned inChapter 1, the stop-and-start nature of urbantransit buses limits the amount of natural gassubstituting diesel, significantly reducing theenvironmental advantage of switching to natu-ral gas.

It is important to bear in mind that emissionlevels are a function of a number of parameters—engine, after-exhaust treatment technology, the“reference” fuel quality (the quality of the fuelto be used in the tests) and driving cycles. En-gines that have been optimized to give lowemissions at steady-state do not necessarily havelow emissions in transient (meaning continuallyvarying speed) driving cycles, which are morerepresentative of real life service. This has ledthe EU to require a new transient test cycle start-ing with the Euro III emissions regulations. Un-like light-duty vehicles where the entire vehicleis driven on a chassis dynamometer for emis-sion certification, heavy-duty vehicles are testedby running engines in engine dynamometers.Transient engine testing is very expensive to carryout, and for this reason the EU and Japan havebeen using only steady-state testing for emis-sions certification of heavy-duty vehicles. In theUnited States, a transient cycle has been usedfor heavy-duty engines since 1985.

The importance of driving cycles was high-lighted in a recent court case in the United States

Table 6. Emissions Benefits of ReplacingDiesel with CNG Vehicles

Fuel CO NOx PM

Diesel 2.4 g/km 21 g/km 0.38 g/kmCNG 0.4 g/km 8.9 g/km 0.012 g/km% reduction 84 58 97

Note: Diesel engines certified to the 1997 U.S. federal emission stan-dards. The numbers are averages of three vehicles in each fuel cat-egory. All were equipped with oxidation catalysts except one CNGbus.Source: Frailey and others 2000.


concerning charges of “cycle beating.” A num-ber of diesel vehicle manufacturers were accusedof carefully optimizing the emissions controlexactly under the conditions that matched thespecified test driving cycle. Outside of the testconditions, the vehicle would be configured tomaximize fuel economy rather than emissionscontrol in real-life driving, resulting in muchhigher emissions. All the engine makers werefined heavily for contravening the spirit, if notthe letter, of U.S. emissions legislation.

Governments in North America and the EUare in the process of introducing much tighteremission standards for vehicles, including heavy-duty diesel vehicles, to be fully enforced by theend of this decade. The so-called Tier 2 emis-sion standards for heavy-duty vehicles, whichwill be phased in beginning with the 2007 modelyear (that is, in the autumn of 2006) in the UnitedStates, will lower particulate and NOx emissionlevels by 90 and 95 percent, respectively. TheEU is introducing progressively tighter standardsin 2005 and 2008. The evolution of standardsfor new heavy-duty diesel vehicles is shown inTable 7. The limits are shown both in grams perkilowatt-hour (g/kWh) and grams per brakehorsepower per hour (g/bhp-h) for the U.S. stan-dards, the latter being the original numbers. Tier2 emission standards in particular are sufficientlystringent that it is not immediately obvious thatNGVs will have significant emissions advantagesover the so-called “clean” diesel.

The clean diesel technology, however, re-quires not only advanced and sophisticated ve-hicle technology and emissions controls, but also

dramatic reductions in diesel sulfur. The needfor what is known as ultra-low-sulfur or sulfur-free diesel arises from the deactivating impactof sulfur on state-of-the-art exhaust control tech-nologies. Compliance with Tier 2 in NorthAmerica requires a sulfur reduction to 15 partsper million by weight (wt ppm). The EU is re-quiring a reduction to 50 wt ppm by 2005, andis seriously considering a 2011 deadline for lim-iting maximum sulfur in diesel to 10 wt ppmsulfur in all diesel. Finland, Sweden and Ger-many are already granting tax incentives to in-troduce diesel with 10 wt ppm. In all developingcountries, the level of sulfur in diesel is muchhigher than that in ultra-low sulfur diesel. Onlya handful of countries, such as Mexico and Thai-land, have a sulfur limit of 500 wt ppm, still alevel too high to meet Tier 2 or Euro IV and Vemission standards. Several developing coun-tries have a diesel sulfur limit as high as 10,000wt ppm (or 1 percent by weight). Deployingclean diesel vehicles would require not onlymanufacturing or importing the best availabletechnology vehicles, but also either importingultra-low sulfur diesel or investing heavily inrefinery modifications to reduce sulfur in dieseldrastically. Massive refinery investment to lowerdiesel sulfur is unlikely to be a high priority inthe near future, especially in countries that havea policy of maintaining a low retail price of diesel.

NGVs may be equipped with an oxidationcatalyst to oxidize (that is, burn completely) re-sidual carbon monoxide and hydrocarbons, suchas lubricant but also methane. Methane is ther-modynamically the most difficult hydrocarbon

Table 7. Heavy-Duty Diesel Emission Standards(in g/kWh, with g/bhp-h in parentheses)

USA USA USA Tier 2 Euro I Euro III Euro IV Euro VPollutant 1990 1998 2007+ 1993 2000a 2005a 2008a

Particulate matter 0.80 (0.6) 0.13 (0.10) 0.013 (0.01) 0.36 0.16 0.03 0.03NOx 8.0 (6.0) 5.4 (4.0) 0.27 (0.20) 8.0 5.0 3.5 2.0Hydrocarbons 1.7 (1.3) 1.7 (1.3) 0.19 (0.14)b 1.1 0.78b 0.552 0.55b

a. European transient cycle.b. Non-methane hydrocarbons.Source: http://www.dieselnet.com/standards.html.


to oxidize and requires a specialized catalyst,but catalyst deactivation has been a problem.Sulfur based odorant (tetrahydrothiophene) at10-15 milligrams per cubic meter (mg/m³) canhave a detrimental effect on oxidation catalystconversion efficiency. Because methane is in-ert, its release is essentially harmless from thepoint of view of local air pollution, but methaneis a powerful greenhouse gas. While U.S. emis-sions regulations distinguish between methaneand non-methane hydrocarbons, those for light-duty vehicles in the EU do not, posing an addi-tional challenge to the manufacturers of NGVs.

As mentioned earlier, particulate emissionsfrom NGVs originate primarily from the lubri-cant; excessive oil consumption could lead tonon-negligible particulate emissions from NGVs.An oxidation catalyst can lower particulate emis-sions some. Although not yet a concern in mostdeveloping country cities, it is worth noting thata NG engine that is not optimized can have muchhigher NOx emissions than a conventional die-sel engine. The reported NOx emission of cur-rent heavy-duty NG engines varies from 0.5 to3.5 g/kWh. Multi-point fuel injection and closed-loop control systems are instrumental in assur-ing low emissions in transient driving.Conversion of Mercedes buses to NG in Brazilis reported to have resulted in higher emissionsof a number of pollutants.

By 2004, on-board diagnostic (OBD) moni-tors will have to be active on all fuels in theUnited States. OBD systems, which monitoremission control components for any malfunc-tion or deterioration that cause an excess inemission limits, alert the driver of the need forrepair via a dashboard light when the diagnos-tic system has detected a problem. In 2005, Eu-ropean OBD compliance will be required onalternative fuels. This could be a major impedi-ment to the growth of alternative fuel vehicles.OBD calibration is very expensive, and eachengine type has to be calibrated. If the numberof vehicles sold is small, the cost of calibrationcannot be economically amortized. Certificationand OBD II (generation two) issues have severely

curtailed aftermarket alternative fuel conversionsin North America.

Stringent Euro IV and V and Tier 2 emissionstandards that may require particulate traps anda (not yet commercially available) “lean deNOx”catalyst system for reducing NOx may possiblywork in favor of CNG vehicles. Depending onfuture technological developments, these factorsmay make natural gas competitive in NorthAmerica and the EU.

Recently, emissions data were obtained onCNG buses (1996, 1998 and 1999 model year)equipped with oxidation catalysts and dieselbuses (1999 model year) fueled by ultra-lowsulfur diesel containing 30 wt ppm sulfur andequipped with Johnson Matthey’s ContinuouslyRegenerating Technology (CRTTM) filter systemfor reducing particulate emissions. The resultsfrom the central business district (CBD) and NewYork (NY) bus cycles are shown in Table 8. CNGbuses had an advantage over diesel for NOx butnot for particulate emissions. The much higherhydrocarbon emissions of CNG buses are dueprimarily to greater methane release. CNG buseswere also found to have higher carbonyl emis-sions. The data in Table 8 should be interpretedwith caution, however, since the measurementswere taken at three different test sites.

It would be useful to reinforce the observa-tion about gasoline vehicle emissions. Absentcatalytic converters, switching from gasoline toa gaseous fuel often meant reduced exhaustemissions. Today, advanced technology gaso-line vehicles with three-way catalysts are so clean

Table 8. Comparison of CNG and “CleanDiesel” Buses (g/km)

CBD cycle NY bus cyclePollutant CNG Diesel CNG Diesel

Particulate matter 0.011 0.015 0.044 0.023NOx 15 16 32 45Total hydrocarbons 10 0.01 42 0.038

Source: 2001, Interim report: Emissions Results from Clean DieselDemonstration Program with CRTTM

Particulate Filter at New YorkCity Transit, http://www.epa.gov/OMS/retrofit/documents/ny_crt_presentation.pdf.


that the fuel itself (that is, whether liquid or gas)plays a minor role, especially for the regulatedemissions. Under these circumstances, convert-ing an advanced gasoline vehicle to gaseous fuelcould even increase rather than decreaseemissions.

FUEL QUALITY

The quality of the fuel affects emissions andvehicle performance. In the case of diesel, theneed to lower sulfur for clean diesel technologyhas been noted. Natural gas quality may varywith the proportion of light hydrocarbons in thegas (especially methane), water, oil and dust.The quality of natural gas in the United States istypically consistently high, with a methane con-tent of 95 or even 98 percent not unusual. Evenso, Sun Metro, the public transportation author-ity of El Paso, Texas, selected LNG for its natu-ral gas-fueled mass transit fleet because thenatural gas quality could not be guaranteed at98 percent methane. In Europe, natural gas qual-ity is more erratic, in particular the methanecontent of U.K. supplies. This is said to be add-ing to the difficulties of meeting future emis-sions standards in the EU.

If a mechanical fuel metering system withoutany kind of feedback control is used, the en-gine has to be tuned for a specific gas. To com-pensate for even small changes in gascomposition, a closed-loop fuel metering sys-tem with an exhaust gas oxygen sensor isneeded.

The NG bus program in Bangkok, Thailand,encountered problems initially on account of themethane content varying from 60 to 85 percent.More specifically, the buses manufactured byMAN with Lambda 1 technology experiencedoperational problems. Today, blending is usedto reduce the amount of gas quality variation.As a result, the variation of the methane contenthas been narrowed to between 72 and 83 per-cent.

INTERNATIONAL EXPERIENCE

The international experience with NG busoperation is limited compared to other typesof vehicles. The world leader is the United States,where 80 transit authorities operated some 3,500NG transit buses in over half of the 50 states asof January 2000 (Watt 2001). Over a quarter ofthe transit authorities had 50 or more NG buses.In Delhi, India, some 1,600 NG buses were inoperation by mid-2001 (CPCB 2001). Anotherleader is China, with over 1,300 NG transit busesin Beijing alone. Other countries with over 200NG transit buses in operation include Australia(492 buses), Canada (367), France (350), Swe-den (320), Japan (259) and Germany (220 in1996). The Republic of Korea has an ambitiousplan to put 5,000 NG transit buses in circulationin eight cities for the 2002 World Soccer Cup.Interestingly, two world leaders in the NGVmarket, Argentina and Italy, have few NG tran-sit buses: Italy has about 170, and Argentina hasnone operating regularly.

Among other developing countries, the UnitedStates Agency for International Development(USAID) has provided funds to purchase a num-ber of NG transit buses to combat air pollutionin Cairo, Egypt. Fifty-seven new dedicated NGbuses operated in Egypt in early 2001. The gov-ernment of Indonesia announced recently thatall new buses and taxis will need to be pow-ered by CNG. The regulations will be introducedfirst in Jakarta and surrounding cities. Most ofJakarta’s 10 million inhabitants rely on buses fortransport. No timetable has been proposed forthe implementation of the regulations. The gov-ernment of Thailand wants to replace 10 per-cent of total oil use in the transport sector withnatural gas in the next five years through a pro-gram estimated to cost 959 million baht (US$23million). The extent to which NG buses will bepromoted is not yet clear.

A detailed description of the international ex-perience with NG transit buses can be found inWatt (2001). Several countries where substan-


tive data are available are discussed brieflybelow.

United States

The United States has the most extensive expe-rience with urban natural gas buses. As of Janu-ary 2000, an estimated 44,300 transit busesoperated in the United States, of which 3,535were reported to be fueled by natural gas, CNGas well as LNG. NG buses accounted for 18 per-cent of new bus orders and 28 percent of po-tential orders. The principal driving force forswitching to NG in the United States is low emis-sions, and hence NG buses are deployed in cit-ies with serious air pollution. All full-size transitbus manufacturers in the United States offer NGbuses. There are 1,200 CNG and close to 70 LNGrefueling stations throughout the country, withover 70 stations serving transit bus fleets.

The federal government heavily subsidizes thepurchase of transit buses, and transit bus fleetsdo not have to rely entirely on local funding.For example, the Federal Transit Administrationsubsidizes up to 83 percent of the cost of a newNG transit bus. Air quality funds offset most ofthe differential capital and some of the infrastruc-ture costs. The federal government supportsNGVs by imposing lower highway tax on CNGand LNG: highway taxes are 6.3 U.S. cents perliter (¢/l) for diesel, 4.7 ¢/l for gasoline, 3.0 ¢/lfor LNG and 1.5 ¢/l of gasoline equivalent forCNG. The federal government also gives grantsof up to US$50,000 per heavy-duty vehicle andup to US$100,000 per refueling station, and al-lows tax deductions for NG vehicle owners. Some32 state governments support NGVs through taxcredits, grants and other incentive schemes. Inone of the most successful CNG/LNG transit busprograms, run by Sun Metro in Texas, the pay-back period for the cost of switching from dieselto NG (including fueling infrastructure) beforegovernment grants is said to be 4.5 years, re-duced to 0.9 years after receiving grants.1

The capital cost of NG transit buses is 10 to25 percent higher than their diesel equivalents,which cost upwards of US$200,000, resulting ina cost increase of up to US$50,000 per bus. Re-fueling, maintenance and bus storage infrastruc-ture is also expensive. Different transit authoritiesreport significantly different incremental oper-ating costs. Two examples illustrate this point.In the first case, two northern California publictransit agencies that replaced old diesel buseswith new CNG buses in 1994 conducted a three-year study. The agencies found that the incre-mental capital cost of bus purchase (excludingthe incremental cost of establishing the CNGinfrastructure) was recovered in about sevenyears (Finley and Daly 1999). In the second case,the Los Angeles County Metropolitan Transpor-tation Authority, which operates the largest fleetof NG buses in the country, reports that the fuelcost for NG buses per distance traveled has been27 percent higher than diesel when compres-sion costs are taken into account and that NGbuses have had a much greater defect rate. Sev-eral other agencies have reported higher oper-ating costs for NG buses; some have decidedto stay with natural gas despite the economicdis incent ives purely on environmentalgrounds. One has concluded that NGVs are asreliable as gasoline vehicles in light-duty applica-tions, but are inferior to diesel in heavy-dutyapplications, and recommended a moratoriumon further expansion of the NG bus programbut continuation with light-duty CNG vehiclepurchase.

Fleet operators found it particularly difficultto recover the incremental costs of mounting aNG bus program in the early days of NG buses,when they experienced many mechanical andother problems. Comments submitted by theAmerican Public Transit Association to the De-partment of Energy in July 1998 highlighted theseconcerns. The key issues to be overcome to in-crease the presence of transit NG buses in themarket were said to be:


� higher capital and startup costs: bus purchase,fueling station, training for maintenance andoperators, garage retrofit

� higher operating costs: fuel station mainte-nance, bus maintenance (lower engine reli-ability, reduced brake life, more expensiveparts due to lower volumes), fuel costs, re-duced range.

Box 1 shows an example calculation illustrat-ing the inter-fuel price difference needed to makeconversion from diesel to CNG economic.

NG bus programs experienced some setbacksin the 1990s, attributed to the cost and reliabilityproblems of the early generations of buses. Somecities reversed their plans to purchase NG buses.There are consistent reports that the generationof NG buses from the early 1990s were not onlymore expensive to purchase than their dieselequivalents, but also were about 30 to 40 per-cent more expensive to maintain and had con-siderably reduced reliability (such as shortermean distance between in-service failures andhigher defect rate with parts).

Recent announcements to purchase CNGbuses include that by the New York City Metro-politan Transportation Authority (MTA), whichplans to expand CNG bus operations between2000 and 2004 by adding 300 more CNG busesto the fleet and converting 2 depots to CNG.During the same period, the MTA also plans toretire all two-stroke engine diesel buses by 2003and adopt clean diesel technology for 3,500buses by retrofitting them with catalyzed par-ticulate trap filters. The MTA’s experience withCNG buses between 1995 and 2000 was not allpositive, however. CNG buses were found to beonly 50 to 75 percent as reliable as diesel buses,although the difference is narrowing. CNG buseswere also found to be 40 percent less energy-efficient in urban service and significantly moreexpensive to operate (Department of Buses2000).

CARB is a strong proponent of NG transitbuses. It is developing a proposal for low-emis-sion transit buses that includes a particulateemission standard of 0.00 (that is to say, lessthan 0.005) g/bhp-h and a NOx standard of 0.1

What is the price difference between diesel andnatural gas needed to make fuel switching eco-nomic for a fleet operator? The following examplegives order-of-magnitude estimates.

The assumptions made in this example are givenbelow.

� There are 100 buses in the fleet.� Each bus travels 80,000 km a year.� The fuel economy falls by 15 percent upon

switching to natural gas.� The refueling station costs US$750,000 to es-

tablish and US$30,000 annually to maintain.� Electricity costs $0.08/kWh.� An additional $350,000 is required for building

modifications.� The incremental cost of NG buses is US$40,000

per bus.

� The cost of natural gas before compression isUS$0.40 per gallon diesel equivalent.

The net present value (NPV) of this NG bus pro-gram was calculated over a time period of 20 yearsat a discount rate of 8 percent. The NPV becamepositive at a diesel price of US$0.88 per gallon, ormore than twice the price of CNG. If the incremen-tal cost of NG bus purchase is assumed to beUS$30,000 per bus, then the break-even diesel pricefalls to US$0.81, still remaining above twice theprice of CNG. In reality, if the life of a bus is shorterthan 20 years (typical life of diesel buses is taken as12-15 years in economic calculations), the break-even price would be even higher.

Source: King and Hutton 2000.

Box 1. An Example of the Economics of CNG Buses in the United States


g/bhp-h by 2008–2012. Originally designed tomandate alternative fuel technology includingnatural gas, the current proposal gives flexibil-ity by not dictating the choice of fuel. However,the current clean diesel technology cannot meetthese emission standards, so that a technologi-cal breakthrough followed by successful com-mercialization will be needed if diesel technology(as opposed to diesel-electric hybrid technol-ogy) is to be used. California’s SCAQMD ap-proved regulations in June 2000 that mandatedfleet operators with more than 15 buses to usenon-petroleum alternative fuels, effectively ban-ning clean-diesel technology. A recent surveyby CARB has shown, however, that outside ofthe SCAQMD, 17 of 22 northern California tran-sit agencies have chosen the clean diesel path,including virtually the entire San Francisco Bayarea.

The experience with transit NG bus opera-tors in the United States seems to suggest thefollowing lessons:

� If the overall operating and maintenance costsare higher for NG buses, there is no hope ofrecovering the incremental cost of bus pur-chase and establishing the NG infrastructure.The operating and maintenance costs can behigher for a combination of reasons—signifi-cantly lower fuel economy for NG, insufficientprice difference between natural gas and die-sel, frequent breakdowns by NG buses, andfewer kilometers traveled on account of in-creased downtime and shorter driving range(making management of bus routes difficultin some cases if buses are refueled only at theirdepots). In a 1999 document, CARB concludesthat even after overcoming these problems toa considerable extent, operating costs of newNG fleets in the future are estimated to beslightly higher than that of new diesel fleets,and the capital costs for NG fleets—initial buspurchase price and the refueling and facilitymodification costs—will continue to be higherthan that for diesel fleets (CARB 1999).

� A large number of buses should be madeto run on natural gas to take advantage ofeconomies of scale, ideally at least a wholedepot.

� Long-term commitment, support and activeinvolvement by management are crucial.

� Financial capability must be in place.� Extensive training of drivers and mechanics

must be undertaken, and qualified and expe-rienced engineers made available to providecompetent support for smooth operation andmaintenance, as well as to ensure safety. Hav-ing trained operators and maintenance staffwho can catch and report problems orchanges in the buses while in operation andduring preventive maintenance is very impor-tant (Box 2).

� Government mandates/regulations alone areinsufficient, and incentives are needed toencourage conversion from diesel to naturalgas. In the United States, grants are availablefrom the federal and state governments toswitch to natural gas.

� The fleet operators should recognize that therewill be additional costs and inconvenienceduring transition. Some have had to strugglewith more frequent breakdowns requiringrepairs for a number of years.

� No significant regulatory hurdles should haveto be overcome.

Australia

The Government of Australia has introduced anumber of measures to support alternative fu-els. The package of federal government programsproviding a strong incentive to switch to alter-native fuels, especially natural gas, include

� CNG Infrastructure Program providing fund-ing up to 50 percent of the cost of installingin excess of 20 public refueling facilities

� Alternative Fuels Conversion Program fund-ing up to 50 percent of the additional cost ofconversion or purchase of new NGVs with a


gross vehicle weight of over 3.5 tons (whichinclude urban transit buses)

� Diesel and Alternative Fuels Grants Schemeensuring that the fuel price advantage of natu-ral gas over diesel is maintained

� Alternative Fuels Grants Scheme applying ex-clusively to urban buses and offering a grantof 12.1 Australian cents per cubic meter ofnatural gas, improving the price advantageof NG over diesel by approximately 10 per-cent.

The response to NG buses in Australia fromfleet owners has been positive on the whole,certainly much more so than in Canada (seebelow). Many achieved economic savings rela-tive to diesel. Emissions advantages of CNGbuses have been demonstrated time and again.Reliability has been a problem, as well as back-firing. Vehicle manufactures have worked closelywith fleet operators and given considerable tech-nical support. The importance of training andeducation of drivers was highlighted when driv-

ers in Sydney, Australia, spoke of lack of accel-eration and poor drivability of natural gas ve-hicles compared to diesel. Comparison trials withdiesel buses showed that the drivers were con-fusing lack of noise from natural gas buses withlack of acceleration. CNG buses are aboutUS$20,000 more expensive than their dieselcounterparts. The Scania NG buses purchasedby Sydney Buses have given a payback periodof about seven years.

Canada

The CNG program in Canada was launched in1983 with a series of economic incentives. Apayback period of about two years was deemednecessary to encourage owners to switch to NG,and the federal government, some provincialgovernments and the natural gas utilities devel-oped incentive packages. The environmentaladvantages of CNG did not become a factor untilabout 1992. The use of NG buses has beenmotivated primarily by consideration for emis-

Box 2. Phoenix Transit

Phoenix Transit in Arizona operates 411 buses, ofwhich 157 are fueled by LNG. The LNG buses are of1998, 1999 and 2000 model year vintages equippedwith catalytic converters. They travel about 80,000km a year, averaging fuel economy of about 0.80km per liter diesel equivalent, compared to dieselbuses achieving 1.3 km per liter. Carbon monoxideand non-methane hydrocarbon emission levels area little lower than for a diesel engine of the sameage; NOx at idle is significantly higher, and NOx un-der load is considerably lower, than diesel. Data onparticulate emissions are not available. There havebeen no significant differences in road call incidentsbetween diesel and LNG buses of the same age.

The views of Phoenix Transit about how theyhave managed the transition to LNG underscorethe importance of training and winning the sup-port of every person in the organization:

“The main challenge is to have trainedoperators and maintenance staff that can

observe and report changes in the buseswhile in operation and during preventivemaintenance. The next challenge is to havebus manufacturers and component manu-facturers working in partnership with theservice and maintenance contractors. Thebus, bus fueling system, refueling system,transmission and engine have all been achallenge but the problems have not keptthe buses out of service and each prob-lem as it occurs is being resolved to makethe bus better.

The bottom line is training, training andmore training. Phoenix Transit initially metresistance from the operators, mechanics,fuelers, and subsequently the union. Theytrained everybody from top managementto bus washers. A little ‘LNG 101’ goes along way, provided you have the answersto conciliate the resistance.”

Source: Watt 2001.


sions improvement. While Canada was active inthe development and use of NG buses in theearly 1990s, the purchase rate declined consid-erably in 1998 and has come virtually to a halt,in sharp contrast to the NG transit bus market inthe United States. The difference between thetwo North American countries is attributed tothe lack of equivalent Canadian and provincialgovernment policies regarding the need to im-prove air quality. Two major NG bus operatorsin Canada—Coast Mountain Bus Company andthe Toronto Transit Commission—are express-ing concerns about the operational problemsand maintenance costs of NG buses and havebeen reported as being inclined to switch outfrom natural gas to other alternative fuels. Thisperception of NG buses is believed to arisefrom the disappointing performance of thebuses purchased in the early 1990s. Some NGbuses in Canada have even been convertedback to diesel.

One unintended consequence of operatingnew NG buses and older diesel buses in paral-lel is the increased use of diesel buses on ac-count of the higher downtime and lowerpassenger carrying capacity of NG buses, asCoast Mountain Bus Company in British Colum-bia has found. Although the company achievedfuel cost savings of 47 percent in 1999, the sav-ings were offset altogether by the 49 percenthigher operating costs.

France

One new bus out of three is fueled by naturalgas in France today. A quarter of all new busesordered are NG buses, selected primarily forenvironmental reasons. The NG buses are “lowfloor” with composite material cylinders locatedin the roof. The buses travel about 40,000 km ayear on urban routes. Most filling stations useslow fill. NG buses cost about US$28,000 toUS$35,000 more than their diesel counterparts.The financial break-even point relative to dieselis generally achieved for fleets of 20 buses or

more. Fleet managers report that they are gen-erally satisfied with NG buses. No significantmaintenance or operational problems have beenencountered. Operation during cold periods hasbeen somewhat problematic, necessitating a waitof 15 to 20 minutes before the engine functionsproperly and upgrades to reduce the number ofbreakdowns. A survey conducted in May 2000found that 90 percent of passengers believedthat NG buses improved air quality, and 96 per-cent stated that NG buses are superior to dieselbuses.

The government lowered taxes on gas for NGbuses between 1998 and 1999 by giving a taxexemption for 24,000 m³ per year per bus. Dur-ing the same period, the market price of dieselincreased, and in addition the government in-creased taxation on diesel. In Poitiers, the break-even point relative to diesel was achieved for afleet operator when the monthly gas consump-tion reached 45,000 m³, or for 16 buses drivingmore than 4,700 km per month.

Transit Bus Industry inDeveloping Countries

Any assessment of fuel switching for transit busesin developing countries must take into accountthe evolving public transport sector in the indi-vidual countries. Transit bus companies in many,if not most, developing countries are cash-strapped. A large number of operators suffer fromfare controls that have made it very difficult toprovide high-quality service. The emergence ofmini-buses in the informal sector—that is, busesin the hands of non-corporate operators, illegalas well as legal—has posed a serious threat tothe survival of transit buses, especially in theformer Soviet Union and Africa. Where theyoperate illegally, these informal sector buses savecosts by minimizing payments to the govern-ment in the form of taxes and license fees. Insome parts of Central Asia, the tax paid by tradi-tional (formal sector) bus operators is estimatedto be an order of magnitude higher than that


paid by informal operators, giving a consider-able advantage to the latter. Because informalsector buses are in the hands of a large numberof owners, even when they operate legally, theymay be more difficult to bring under control forthe purpose of monitoring and enforcing regu-lations (such as vehicle registration, and safetyand emission standards).

As governments attempt to reduce emissionsfrom buses, they must face that traditional busoperators are cash-strapped, in part because offare controls, and have little money left to main-tain their vehicles properly, and that bus opera-tors in the informal sector are difficult to regulate.In part because they are cash-strapped, bus op-erators do not maintain vehicles, resulting in highemissions as manifested by black smoke belch-ing out of diesel buses. For the same reason,they would not be in a position to purchasemore expensive NG buses, provide extensivestaff training on this new technology, and ac-cept the possibility of more repairs to deal withgreater frequency of bus breakdowns, at leastinitially. That even U.S. bus operators have facedconsiderable mechanical challenges with NGbuses, resulting in higher operating costs in anumber of cases, is a cause for concern, espe-cially given the much smaller number of techni-cally qualified mechanics to service NGVs indeveloping countries. In the informal sector, busoperators are not likely to own a large numberof vehicles. Even if they are operating legally,they are not likely to be able to exploit econo-mies of scale in maintenance, training and refu-eling, making it difficult to switch to NG.

Under the current circumstances, the transitbus industry in a number of countries is not sus-tainable in the long run—fare control will even-tually have to be lifted to bring in new types ofservice, or else the formal sector may disappearaltogether and be replaced by the informal sec-tor. That is to say, high emissions from dieselbuses are not merely because of the choice offuel, but are symptomatic of deeper problems

in the transit bus industry in developing coun-tries, and these same problems may condemnNG bus programs to failure even if inter-fuelpricing is adjusted to favor NG much more atthe expense of diesel. While mini-buses have aproper role to play, large transit buses are idealfor segregated busways in congested situationswhere the reservation of well-functioning rights-of-way for buses is often the only affordablesolution for mass transit. They are also the roadvehicle of choice where passenger volumes arehigh and public transport vehicles constitute ahigh percentage of traffic in congested or near-congested streets.

Some governments, such as the United Statesand Australia, have offered considerable finan-cial incentive packages to promote NG transitbuses. Especially when these incentive packageshave been combined with tough emissions regu-lations, NG bus programs have been successful.However, the level of subsidies offered in thesecountries are unlikely to be sustainable in de-veloping countries.

Experience in developing countries with NGtransit buses is limited. The extent to which ex-perience and lessons—especially with respectto mechanical reliability, drivability, maintenanceand other issues—from industrial countries canbe transferred is not clear. The NG buses nowbeing purchased in industrial countries are allOEM buses, costing some US$20,000 toUS$50,000 more than their diesel counterparts,up to as much as US$300,000 per bus. Most de-veloping countries pay US$100,000 or less foreach bus. Reliability and increased repair fre-quency have been one of the major issues inthe past, even with OEM NG buses. The dataneeded to establish whether NG buses offeredat dramatically lower prices in developing coun-tries such as China have comparable, more, orfewer operational and maintenance problemswill become available only in a few years’ time.

Finally, there is the question of what hap-pens to NG buses if they are neglected as much


as conventional diesel buses in developingcountries. This is an interesting question forcities such as Delhi, which has mandated con-version from diesel to NG. How the poor cul-tural acceptance of, as well as inability to payfor, regular maintenance affects the life of theNG bus, its fuel economy, frequency of com-plete breakdown (so that it cannot be oper-ated on the road), and emissions is animportant question that should be monitored

closely as developing country cities launch NGbus programs.

NOTE

1. However, this analysis by Sun Metro has been

criticized for not comparing like with like, and there-

fore projecting more favorable economics for NG

buses (Watt 2001).

41

Chapter 4

Looking to the Future

T his report has shown that NGVs are arelatively new and rapidly evolving tech-nology. As such, while there are useful

lessons to be learned from other countries’ ex-perience, some of them may no longer be di-rectly applicable—for example, experiencerelating to the poor performance of NG busesmanufactured in the early 1990s compared tothose manufactured more recently.

To date, NG buses have been at a privateeconomic disadvantage compared with dieselbuses unless supported by substantial favorabletax discrimination or subsidies. In the absenceof emissions standards that effectively requiregaseous fuels, natural gas buses are unlikely tobe adopted because they are more expensive tooperate relative to diesel buses. This is partlybecause diesel is a very cheap fuel in most de-veloping countries—it is lightly taxed or mayeven be subsidized. Even if diesel were taxedmuch more, however, it is not obvious that CNGbuses would be cheaper over their life cyclethan diesel buses: they cost more to purchase,are less fuel efficient, have a smaller range andare often less reliable. These observations sug-gest that the social case for replacing diesel byCNG in buses rests on environmental grounds.In particular, the use of natural gas by heavy-duty vehicles normally fueled by diesel wouldnot be suitable if the diversification of energysources is the primary objective.

Developments in the clean diesel technologyare expected to have a direct impact on the fu-

ture of the NGV industry and market in indus-trial countries. The environmental concerns di-recting research and development in the autoindustry in industrial countries are not the sameas those in developing countries. Severe NOx

emissions control, which is not yet a priority inmost developing country cities, presents a sig-nificant technical challenge to vehicle manufac-turers in industrial countries who have turnedto sophisticated technical solutions based onultra-low sulfur diesel (preferably below 10 wtppm sulfur). For controlling particulate emis-sions, catalyzed particulate traps appear to besuccessful in reducing emissions dramatically,but they too require ultra-low sulfur diesel.Cummins cited quirks of Euro III emission stan-dards (including the limit on exhaust methane)and subsequent certification test cycles as theprimary reasons for its earlier decision to with-draw its natural gas engines from the Europeanmarket when Euro III truck and bus emissionslegislation was originally scheduled to come intoforce for all new vehicles in October 20011 . Thatis to say, the evolution of the NGV market andmanufacturing industry in industrial countries isinfluenced to a significant extent by consider-ations that are often not priority issues in devel-oping countries. And yet because the bulk ofproduct development occurs in industrial coun-tries, what happens there will have an impacton the availability of NGVs in developing coun-tries in the foreseeable future. To take an ex-treme scenario, if the NGV industry in industrial


countries were to die (for example, as a resultof clean diesel replacing natural gas), it couldbecome much more difficult for developingcountry cities to implement a NG bus program.

A closely related factor is developments inthe refining industry. More specifically, giventhe future mandated reductions in sulfur in die-sel, there has been extensive research and de-velopment to lower the cost of dieselhydrodesulfurization technologies. Break-throughs have been announced and demon-strated at the pilot scale, including theannouncement of a novel sulfur extraction tech-nology that could radically reduce the cost ofboth diesel and gasoline desulfurization whileefficiently saturating aromatics and boosting ce-tane—the cost of virtually eliminating sulfur inhigh-sulfur diesel is estimated to be on the or-der of 0.7 U.S. cents per liter (Hart’s Diesel FuelNews 2001). Successful commercialization ofsuch processing technologies could dramaticallyalter the landscape for the clean diesel–naturalgas debate.

However, until such a time as “cheap” cleandiesel becomes widely available worldwide,which is not expected for at least several moreyears, most developing country cities will con-tinue to grapple with a choice between conven-tional, polluting diesel versus potentially muchcleaner natural gas buses. If the government ofa city decides that the reduction in air pollutionassociated with CNG buses is worth the cost,then it needs to adopt policies that would en-courage the switch to CNG: either emissions stan-dards for buses, or fuel or vehicle taxes thatreflect marginal social costs.

The type of analysis that could be carried outto determine whether CNG buses should replaceconventional diesel buses would include

� comparing the lifecycle costs and emissionsof a new conventional diesel bus and a CNGbus, using a net of tax price for diesel fuel,CNG and their respective vehicles, or

� comparing the cost of retrofitting a diesel buswith a CNG engine and estimating the result-

ing change in emissions over the remaininglife of the bus (also using net of tax fuel andvehicle prices).

The first is applicable to the case where anew bus purchase is being considered, and thesecond to the case of converting existing dieselbuses (which is not normally recommended).These calculations would give an idea of howthe options compare on economic grounds.Repeating the above calculations using gross oftax fuel and vehicle prices would indicate howmuch more fleet operators would have to bearto achieve target emission levels. By makingassumptions about exhaust emission factors, itwould also be possible to compute a cost perton of particulate matter reduced, although thismight be misleading in view of the fact that otherpollutants are being reduced at the same time.

Many unknowns in the above calculations in-troduce large uncertainties in the final results.The relative operating costs of diesel and NGbuses have been reported to vary over a widerange, even in the United States where data col-lection has been rigorous. Little is known aboutthe experience of converting existing dieselbuses to CNG, other than that such conversionstypically do not make happy customers. Thereliability of CNG buses in developing countriesis one of the greatest unknowns. However, car-rying out calculations using the most optimisticas well as more pessimistic assumptions couldgive order-of-magnitude estimates. For example,if even the most optimistic of assumptions can-not justify CNG buses, then switching to CNG isunlikely to be sustainable.

An added consideration is the cash-strappedstate of the transit bus industry in developingcountries. Lack of an adequate operating bud-get is one of the reasons for the poor mainte-nance of diesel buses, resulting in grossemissions. Since operating NG buses incurshigher upfront costs—for purchasing buses, set-ting up refueling stations and training mechan-ics and drivers—transit bus fleet operators inpoor financial condition are not in a position to

43Looking to the Future

take on a NV bus program successfully. Whilelack of proper maintenance may not necessarilylead to black smoke emitted by tailpipes in thecase of NG buses, it could easily result in morefrequent breakdowns and other operationalproblems, as well as higher emissions of otherpollutants that may become an issue in the longrun. NG buses tend to be less reliable than theirdiesel equivalents even in the best of circum-stances. Reports of continual operational prob-lems with NG buses are certain to invite abacklash from bus operators, seriously harmingthe future of the NGV industry. Some of the morenegative experiences in developing countriesinclude that in Jakarta, Indonesia, where of the40 dedicated CNG buses, only 20 are operatingowing to maintenance problems.

Lastly, because natural gas buses have higherupfront costs that require some type of govern-ment support (for example, in the form of higherdiesel fuel or vehicle taxes) to recover, and henceemissions reductions from switching to naturalgas come at a price, this cost for reducing par-ticulate emissions should be compared to thatfor other sources—for example, industry, house-holds and the informal sector. In order to makethis comparison, the contributions of varioussources to ambient concentrations have to beunderstood. While it is difficult to identify sourcesaccurately, chemical analysis of particles andother analytical studies go a long way in provid-ing a better understanding of source contribu-tions. At the same time, such numerical findingsneed to be tempered by the growing evidencethat diesel particulate emissions are indeed verytoxic, and exhaust emissions fall predominantlyin the particle size range that seems to have themost health impact (namely below 1 micron).

In summary, the following are some of thequestions that should be posed in consideringthe choice of fuel for transit buses.

� What is the financial position of the transitbus operators? If they are cash-strapped, isthis because distortions in the policy frame-work need to be corrected? If fare control is

the primary reason for the bus operators’ poorfinancial state, how much would the faresneed to rise before the financial health of theoperators is recovered? If fares are to be raised,are there provisions to protect low-incomebus riders who may not have alternatives?

� Is automotive diesel priced much above natu-ral gas for transport? If not, switching to natu-ral gas will not be economic, and hence notcommercially sustainable in the long run. Arethere pricing distortions today that may becorrected in the future, and that may have anadverse impact on relative prices of the twofuels (for example, a heavy subsidy for natu-ral gas)? If so, economic calculations shouldbe based on long-run marginal cost of natu-ral gas rather than the current low price. Isthe price difference between natural gas anddiesel (possibly combined with higher vehicletax on diesel vehicles) adequate for recover-ing the incremental costs? Is the payback pe-riod reasonable? Does the government havea plan to support the price difference in timesof falling international price of diesel and ris-ing price of natural gas?

� Is emissions reduction in transport, and morespecifically targeting buses, likely to be cost-effective compared to emissions reductions inother sectors? If, on the contrary, informalrefuse burning, combustion of biomass in ur-ban households, wood and coal burning incottage industries in the informal sector, andtwo-stroke engine gasoline motorcycles andthree-wheelers turn out to contribute themajority of ambient concentrations of particu-late matter, targeting diesel buses aggressivelywill not reduce particulate air pollutionmarkedly.

� How will large-scale substitution away fromgasoline to NG affect the government’s fi-nances? Assuming the price of diesel is be-low the price of gasoline, setting the taxrate on NG that makes it economically at-tractive to switch from diesel to NG willmake it even more attractive to switch fromgasoline to NG. Will a successful NGV pro-


gram require adjustments to inter-fuel taxa-tion later on, making NG less attractive? Willthere be serious imbalances in petroleumproduct consumption patterns, such as a verylow gasoline-to-diesel ratio? Such imbalancesmay not be a problem for a country that im-ports nearly all of its fuel demand, but theywould constrain the ability of refineries toremain profitable.

� Are subsidies (such as capital subsidies pro-vided by the U.S. government) needed to jus-tify conversion to natural gas on economicgrounds? If diesel tax is not raised, shouldCNG buses even be considered? Is there roomfor subsidies as the United States has pro-vided for capital? If so, based on order-of-magnitude calculations, does it appear thatthe benefits of switching to natural gas in termsof health impacts justify the subsidies? Or canthe government save more lives and reduceillnesses by using the same amount of moneyelsewhere, such as on clean water or healthcare? If subsidies seem justified, is the gov-ernment committed to providing subsidies inthe long run to avoid the repeat of the NewZealand NGV experience?

� Is the regulatory framework for NGVs in place,including safety regulations and standards forequipment? Is there an adequate monitoringand enforcement mechanism?

� Is the quality of natural gas consistent anddoes it meet minimal requirements for vehicles,or does it tend to fluctuate over a wide rangeso that certain vehicle technologies may notbe used without processing the gas further(such as by blending)?

� How many buses could be converted to natu-ral gas at one depot? Is there a sufficient num-ber to exploit economies of scale?

� Is there a realistic and workable plan to trainmaintenance staff and drivers? Are thereenough qualified trainers and engineers tosupport the program?

� Are the transit fleet managers considering NGbuses very much committed to the fuel switchand ready to get involved themselves?

The decision to switch from diesel to naturalgas for use in buses is not straightforward. At aminimum, the regulatory and administrative ar-rangements should be in place to ensure thefinancial sustainability of transit operators whowould be using natural gas, and vehicle taxesshould reflect marginal social costs of healthdamage from air pollution. If these conditionsare satisfied, cities in countries with abundantsupplies of domestic gas, gas pipelines alreadyin place, and with transport emissions contrib-uting substantially to serious urban air pollutionmay consider this fuel option. The conditionsoutlined above would present a serious chal-lenge even in industrial country cities, and cer-tainly in nearly all developing country cities.

The temptation to mandate NG buses underthe circumstances may be strong, but mandat-ing what would otherwise be commercially un-sustainable in the long run cannot be a viablesolution. If municipal governments are seriousabout wanting to promote NG for transit buses,broader issues facing the public transport sec-tor, such as examination of the fare structureand how to create a level playing field for allbus operators including regulation of informalsector buses, will need to be addressed. Oncetransit bus operators are on their way to a moresound financial footing, the question of how bestto recover the incremental cost of switching toNG—such as fiscal incentives reserved only forfleet operators so as not to increase the fiscalburden on the government—may be explored.Lastly, as a growing number of developing coun-try cities experiment with NG buses, systematiccollection and exchange of information couldbe invaluable in guiding policymakers as theyconsider this new fuel system.

NOTE

1. In January 2001, the EU decided to delay imple-mentation of Euro III rules for gas engines by twoyears. In response, Cummins has put on hold its de-cision to withdraw from the European gas enginemarket.

45

Annex A

Emissions from Diesel Vehicles

T he pollutants of concern found indiesel exhaust are particulate matter,oxides of nitrogen (NOx) and hydro-

carbon toxins such as polynuclear aromatics.Primary particles are emitted directly by vehicles;secondary particles, in contrast, are formed fromthe chemical oxidation of atmospheric gases.Oxides of sulfur (SOx) and of nitrogen (NOx)are precursors for secondary particles; reducingsulfur in diesel lowers the amount of sulfate-based particles. NOx is in addition an ozone pre-cursor. Ozone pollution is not yet a seriousproblem in most developing country cities, butambient concentrations of ozone and NOx areon the rise.

There is a growing consensus that diesel ex-haust poses a cancer risk. The advisory board tothe U.S. National Toxicology Program has rec-ommended that diesel exhaust particles be listedas “reasonably anticipated to be a human car-cinogen.” The California Air Resources Board(CARB) has officially recognized that some ele-ments of emissions from diesel engines are car-cinogens. CARB, in fact, points to several studiesthat have shown that the cancer risk from dieselparticles is greater than the risk from all otheridentified toxic air contaminants combined. Japa-nese scientists claim that they have found 3-nitrobenzanthrone to be one of the mostcarcinogenic substances ever discovered; emis-sions of 3-nitrobenzanthrone increase markedlywhen a diesel engine is operating under high

load. This suggests that diesel particulate emis-sions are especially harmful to public health—amatter for concern, since the consumption ofdiesel far exceeds that of gasoline in many de-veloping countries.

A series of extensive studies, mainly in theUnited States, has shown clear associations be-tween small changes in a wide range of healthindicators—mortality, hospital admissions, emer-gency room visits, time off school or work, res-piratory symptoms, exacerbation of asthma andchanges in lung function—and ambient particu-late concentrations. Of the various health indi-cators, the measurement of mortality has beenparticularly well studied. The actual adverseimpact of fine particulate matter on public healthmay be considerably greater in developing coun-tries than existing data indicate: most studies havebeen carried out on urban populations in indus-trial countries who receive high-quality medicalcare and who do not spend as much time out-doors as some segments of the population indeveloping countries do.

Historically, carbonaceous contributions todiesel particulate emissions have far exceededsulfate contributions. Figure A1 shows the re-duction in the carbon soot and organics portionof particulate matter that was achieved by im-proving heavy-duty diesel engine design be-tween 1988 and 1994 in the United States. Thesulfate portion is due to sulfur found in dieseland lubricant. In 1988, carbon soot and organics


constituted the majority of particulate emissions.As a result of steady refinements in the enginetechnology, the sulfate contribution was greaterthan the carbonaceous contribution by 1994. Theengine manufacturers would not have been ableto meet the 1994 particulate emission limit of0.1 grams per brakehorse power per hour (g/bhp-hr) without reducing sulfur in diesel from0.25 percent to 0.05 percent. Therefore, the de-cision of the U.S. Environmental ProtectionAgency (EPA) to lower the limit on sulfur in die-sel to 0.05 percent for the 1994 model year wasjustified and should serve as a model for futureregulatory action—vehicle technology and fuelquality improvements should be consistent andcoordinated.

Certain fuel parameters have been linked todiesel emissions. One of the most extensive stud-ies conducted to date is the EuropeanProgramme on Emissions, Fuels and EngineTechnologies (EPEFE). Representing an unprec-edented collaboration between the Europeanmotor and oil industries, EPEFE was undertakenas part of the European auto-oil program, the

objective of which was to identify measures forimproving urban air quality based on sound sci-ence and cost-effectiveness as primary criteria.EPEFE found that the relationships between ve-hicle and fuel technologies were complex, sothat those measures that reduced emissions fromlight-duty vehicles sometimes increased emis-sions from heavy-duty vehicles, and vice versa.For example, reducing diesel density decreasedNOx emissions from heavy-duty diesel engines,but increased NOx from light-duty engines. Somemeasures were effective for all diesel enginestested: decreasing polynuclear aromatics in die-sel reduced NOx and particulate emissions, andincreasing cetane number decreased hydrocar-bon and carbon monoxide emissions, from bothheavy- and light-duty vehicles. However, increas-ing cetane number increased particulate emis-sions from light-duty vehicles. Overall, thefollowing measures were found to decrease par-ticulate emissions:

� reducing density for light-duty engines� reducing polynuclear aromatics for both light-

and heavy-duty engines� decreasing the temperature at which 95

percent of diesel evaporates, for light-dutyengines.

The health impact of diesel emissions appearsto be especially serious for those close to thesources of emissions, such as school childrenriding buses, traffic police and vehicle ridersfollowing diesel vehicles. As a striking example,one study in Los Angeles, United States, foundthat concentrations of elemental carbon (whichconstitutes a relatively high fraction of dieselparticulate emissions) inside vehicles with win-dows closed were about 5 micrograms per cu-bic meters (µg/m³) without any vehicles in front;15 µg/m³ when following a diesel truck with ahigh, vertical exhaust pipe or a diesel passengercar; 50 µg/m³ when following a diesel truck witha low exhaust pipe; and as high as 130 µg/m³when following an urban transit bus making fre-quent stops (Fruin and others 2000). These con-

Figure A1. Particulate Emissions from NewVehicles in the United States

Source: McCarthy 1994.

1988 1991 1994 19940.25%sulfur

0.25%sulfur

0.25%sulfur

0.25%sulfur

g/bhp-hr

Carbon soot and organics

Sulfate

0.6

0.5

0.4

0.3

0.2

0.1

0

47Annex A: Emissions from Diesel Vehicles

centrations would be expected to be even higherin developing countries where emission levelsare generally higher than in California.

While there are growing concerns about theenvironmental health risks of conventional die-sel vehicles as studies continue to shed new lighton the adverse health impact of particulate emis-sions, the markedly higher efficiency of dieselengines compared to gasoline has made dieselpopular. For example, in Europe, one out ofevery three cars sold in 2001 is forecast to bediesel-fueled. The EU and the United States havesteadily tightened emission standards, requiringengine modifications as well as diesel qualityimprovements. The challenge facing vehiclemanufacturers is the trade-off between NOx andparticulate emissions: measures that reduce thepeak flame temperature, such as injection tim-ing retard, decrease NOx emissions but increaseparticulate emissions that arise from incompletecombustion, and hence fuel consumption.

Oxidation catalysts are used to reduce gas-eous hydrocarbons and the soluble organic frac-tion of particles in the diesel exhaust. Oxidationcatalysts are especially effective for two-strokeengine diesel vehicles. In the Urban Bus Retro-fit/Rebuild Program finalized by the U.S. EPA in1993, oxidation catalysts are extensively used toreduce particulate emissions. The program is lim-ited to 1993 and earlier model-year urban busesoperating in metropolitan areas and applies atthe time of engine rebuild or replacement. Akey aspect of the program is the certification ofretrofit/rebuilt equipment, which often includesan oxidation catalyst as one of the components.The use of oxidation catalysts calls for low (butnot necessarily ultra-low) sulfur diesel, namelydiesel with a sulfur content of 0.05 percent orlower.

For dramatic reductions of particulate emis-sions, catalyzed particulate filters and continu-ously regenerating particulate traps have beenshown to be effective. A continuously regener-ating particulate trap can reduce particle num-ber counts by one to two orders of magnitudeas well as the mass of particles. To use theseparticulate filters, however, ultra-low sulfur isneeded for durability. In response, the EuropeanCommission proposed during the first half of2001 that sulfur-free diesel (meaning diesel withsulfur below 0.001 percent by weight, or 10 partsper million) be made available in all EU coun-tries by 2005.

In the United States, in-use diesel vehicles aretested for smoke opacity in eight states. The U.S.EPA has not mandated opacity tests for dieselvehicles because of concerns about correlationbetween opacity measurements and particulateemissions. While loaded dynamometer tests arereasonably correlated with mass particulate emis-sions to a degree, lowering smoke opacity doesnot necessarily guarantee a reduction in particu-late emissions, and vice versa. The correlationbetween opacity measurements under snap ac-celeration1 (which is the test used even in theUnited States) on one hand and mass particu-late emissions during transient operation is muchweaker. This poses a considerable challenge fordesigning an effective diesel vehicle inspectionand maintenance program.

NOTE

1. The vehicle engine, with the transmission in

neutral, is accelerated at full throttle from a raised

idle revolutions per minute (rpm) to a maximum gov-

erned rpm.

49

References

Airborne Particles Expert Group. 1999. “SourceApportionment of Airborne Particulate Matterin the United Kingdom,” http://www.aeat.co.uk/netcen/airqual/reports/home.html#reports

Automotive Environment Analyst. 2001. “CNGcar explodes,” No. 76, May.

bp. 2001. bp Statistical Review of World EnergyJune 2001, June, http://www.bp.com/centres/energy/index.asp

Bacon, Robert W. 2001. “Oil Product Taxes.”Forthcoming in Viewpoint, World Bank, Wash-ington, D.C.

CARB (California Air Resources Board). 1999.“Proposal for Cleaner Transit Buses,” http://www.arb.ca.gov/msprog/mailouts/msc9928/msc9928a.doc

CARB (California Air Resources Board). 2001.“Public Meeting to Discuss Motor Vehicle CNGFuel Specifications,” http://www.arb.ca.gov/fuels/altfuels/meeting/2001/0307arbp.pdf

CPCB (Central Pollution Control Board). 2001.“Vehicular Pollution Control in Delhi: Initia-tives and Impacts.” August, Delhi, India.

Cumming, Robert. 1997. “Vehiculos A Gas Natu-ral: Estado Actual Y Perspectivas,” presenta-tion made at the Seminario Internacional,Mexico City, 11-12 September.

Department of Buses. 2000. “NYTC Clean FuelBus Programs,” presentation made at theWashington Metropolitan Area Transit Author-ity Alternative Fuels Workshop, 6 July 2000,

http://www.epa.gov/OMS/retrofit/docu-ments/nyc_2.pdf

Finley, Bruce E. and Tracy A. Daly. 1999. “AThree Year Comparison of Natural Gas and Die-sel Transit Buses,” SAE Technical Paper Series,1999-01-3738. Warrendale, Pennsylvania.

Frailey, Mike, Paul Norton, Nigel N. Clark andDonald W. Lyons. 2000. “An Evaluation ofNatural Gas versus Diesel in Medium-DutyBuses,” SAE Technical Paper Series, 2000-01-2822. Warrendale, Pennsylvania.

Francchia, Juan Carlos. 2000. “An Overview ofthe Argentine NGV Experience,” presentationmade at the Workshop on Compressed Natu-ral Gas, 2-3 March, Washington, D.C.

Fruin, S.A., S.P. Hui, P.L Jenkins and C. Rodes.2000. “Fine Particle and Black Carbon Con-centrations inside Vehicles,” presentationmade at the 10th Annual Conference of theInternational Society of Exposure Analysis,Monterey, California, October 25.

Harris, Garth. 2000. “Compressed Natural Gasin New Zealand,” presentation made at theWorkshop on Compressed Natural Gas, 2-3March, Washington, D.C.

Hart’s Diesel Fuel News. 2001. “Sinclair/Bechtel/SulphCo Evaluation Indicates Huge Cost Re-duction for Fuels Desulfurization,” Vol. 5, No.6, March 19.

Impco Technologies. 2000. “Alternative FuelsPresentation to the World Bank,” presenta-


tion made at the Workshop on CompressedNatural Gas, 2-3 March, Washington, D.C.

King, Thomas A. and Mark B. Hutton. 2000. “Eco-nomic, Environmental, and Technical Aspectsof Public Transit NGVs,” presentation madeat the Workshop on Compressed Natural Gas,2-3 March, Washington, D.C.

Maxwell, Timothy T. and Jones, Jesse C. 1995.Ch. 5, “Conversion of Compression IgnitionEngines.” In Alternative Fuels: Emissions, Eco-nomics and Performance. Society of Automo-tive Engineers, Inc., Warrendale, Pennsylvania.

McCarthy, Christopher I. 1994. “Update on theEffect of Government Regulations on DieselFuels,” presentation made at the 1994 National

Petroleum Refiners Association National Fu-els and Lubricants Meeting, 3-4 November,Houston, Texas.

Nylund, Nils-Olof and Alex Lawson. 2000. “Ex-haust Emissions from Natural Gas Vehicles:Issues related to engine performance, exhaustemissions and environmental impacts,” reportprepared for the IANGV Technical Commit-tee, 31 March, http://www.iangv.org/html/sources/sources/reports/emissions.html

Watt, Glen M. 2001. “Natural Gas Vehicle TransitBus Fleets: The Current International Experi-ence,” IANGV Review Paper, http://www.iangv.org/html/sources/sources/re-ports/iangv_bus_report.pdf

Documents

Breathing Clean - cti2000.it World Bank Methane buses.pdf · gas vehicle fleet in the world, has in fact no natural gas buses in regular operation. In devel-oping countries, diesel