5.3.1.1 Reducing vehicle loads

Lightweight materials

A 10% weight reduction from a total vehicle weight can improve fuel economy by 4–8%, depending on changes in vehicle size and whether or not the engine is downsized. There are several ways to reduce vehicle weight; including switching to high strength steels (HSS), replacing steel by lighter materials such as Al, Mg and plastics, evolution of lighter design concepts and forming technologies. The amount of lighter materials in vehicles has been progressively increasing over time, although not always resulting in weight reductions and better fuel economy if they are used to increase the size or performance of the vehicle. In fact, the average weight of a vehicle in the USA and Japan has increased by 10–20% in the last 10 years ( JAMA, 2002 [NPR] ; Haight, 2003 [NPR] ), partly due to increased concern for safety and customers’ desire for greater comfort.

Steel is still the main material used in vehicles, currently averaging 70% of kerb weight. Aluminium usage has grown to roughly 100 kg per average passenger car, mainly in the engine, drive train and chassis in the form of castings and forgings. Aluminium is twice as strong as an equal weight of steel, allowing the designer to provide strong, yet lightweight structures. Aluminium use in body structures is limited, but there are a few commercial vehicles with all Al bodies (e.g., Audi’s A2 and A8). Where more than 200 kg of Al is used and secondary weight reductions are gained by down-sizing the engine and suspension – more than 11–13% weight reduction can be achieved. Ford’s P 2000 concept car ^[11] has demonstrated that up to 300 kg of Al can be used in a 900 kg vehicle.

Magnesium has a density of 1.7–1.8 g/cc ^[12] , about 1/4 that of steel, while attaining a similar (volumetric) strength. Major hurdles for automobile application of magnesium are its high cost and performances issues such as low creep strength and contact corrosion susceptibility. At present, the use of magnesium in vehicle is limited to only 0.1–0.3% of the whole weight. However, its usage in North American-built family vehicles has been expanding by 10 to 14% annually in recent years. Aluminium has grown at 4–6%; plastics by 1–1.8%; and high strength steels by 3.5–4%. Since the amount of energy required to produce Mg and also Al is large compared with steel, LCA analysis is important in evaluating these materials’ potential for CO₂ emission reduction ( Helms and Lambrecht, 2006 [NPR] ). Also, the extent of recycling is an important issue for these metals.

The use of plastics in vehicles has increased to about 8% of total vehicle weight, which corresponds to 100-120 kg per vehicle. The growth rate of plastics content has been decreasing in recent years however, probably due to concerns about recycling, given that most of the plastic goes to the automobile shredder residue (ASR) at the end of vehicle life. Fibre-reinforced plastic (FRP) is now widely used in aviation, but its application to automobiles is limited due to its high cost and long processing time. However, its weight reduction potential is very high, maybe as much as 60%. Examples of FRP structures manufactured using RTM (resin transfer method) technology are wheel housings or entire floor assemblies. For a compact-size car, this would make it possible to reduce the weight; of a floor assembly (including wheel housings) by 60%, or 22 kg per car compared to a steel floor assembly. Research examples of plastics use in the chassis are leaf or coil springs manufactured from fibre composite plastic. Weight reduction potentials of up to 63% have been achieved in demonstrators using glass and/or carbon fibre structures ( (Friedricht, 2002 ) ).

Aside from the effect of the growing use of non-steel materials, the reduction in the average weight of steel in a car is driven by the growing shift from conventional steels to high strength steels (HSS). There are various types of HSS, from relatively low strength grade (around 400 MPa) such as solution-hardened and precipitation-hardened HSS to very high strength grade (980– 1400 MPa) such as TRIP steel and tempered martensitic HSS. At present, the average usage per vehicle of HSS is 160 kg (11% of whole weight) in the USA and 75 kg (7%) in Japan. In the latest Mercedes A-class vehicle, HSS comprises 67% of body structure weight. The international ULSAB-AVC project (Ultra Light Steel Auto Body – Advanced Vehicle Concept) investigated intensive use of HSS, including advanced HSS, and demonstrated that using HSS as much as possible can reduce vehicle weight by 214 kg (–19%) and 472 kg (–32%) for small and medium passenger cars respectively. In this concept, the total usage of HSS in body and closures structures is 280–330 kg, of which over 80% is advanced HSS (Nippon ( Steel, 2002 ) ).

Since heavy-duty vehicles such as articulated trucks are much heavier than passenger vehicles, their weight reduction potential is much larger. It is possible to reduce the weight of tractor and trailer combination by more than 3000 kg by replacing steel with aluminium ( EAA, 2001 [NPR, MoS] ).

Aerodynamics improvement

Improvements have been made in the aerodynamic performance of vehicles over the past decade, but substantial additional improvements are possible. Improvement in aerodynamic performance offers important gains for vehicles operating at higher speeds, e.g., long-distance trucks and light-duty vehicles operating outside congested urban areas. For example a 10% reduction in the coefficient of drag (C_D) of a medium sized passenger car would yield only about a 1% reduction in average vehicle forces on the US city cycle (with 31.4 km/h average speed), whereas the same drag reduction on the US highway cycle, with average speed of 77.2 km/h, would yield about a 4% reduction in average forces. ^[13] These reductions in vehicle forces translate reasonably well into similar reductions in fuel consumption for most vehicles, but variations in engine efficiency with vehicle force may negate some of the benefit from drag reduction unless engine power and gearing are adjusted to take full advantage of the reduction.

For light-duty vehicles, styling and functional requirements (especially for light-duty trucks) may limit the scope of improvement. However, some vehicles introduced within the past five years demonstrate that improvement potential still remains for the fleet. The Lexus 430, a conservatively styled sedan, attains a C_D (coefficient of aerodynamic drag) of 0.26 versus a fleet average of over 0.3 for the US passenger car fleet. Other fleet-leading examples are:

• Toyota Prius, Mercedes E-class sedans, 0.26
• Volkswagen Passat, Mercedes C240, BMW 320i, 0.27

For light trucks, General Motors’ 2005 truck fleet has reduced average C_D by 5–7% by sealing unnecessary holes in the front of the vehicles, lowering their air dams, smoothing their undersides and so forth (SAE International, 2004 ).

The current generation of heavy-duty trucks in the United States has average C_Ds ranging from 0.55 for tractor-trailers to 0.65 for tractor-tandem trailers. These trucks generally have spoilers at the top of their cabs to reduce air drag, but substantial further improvements are available. C_D reductions of about 0.15, or 25% or so (worth about 12% reduced fuel consumption at a steady 65 mph ^[14] ), can be obtained with a package of base flaps (simple flat plates mounted on the edges of the back end of a trailer) and side skirts ( McCallen et al., 2004 [NPR] ). The US Department of Energy’s 2012 research goals for heavy-duty trucks USDOE, research goals for heavy-duty trucks USDOE,) ^[15] include a 20% reduction (from a 2002 baseline, with C_D of 0.625) in aerodynamic drag for a ‘class 8’ tractor-trailer combination. ^[16] C_D reductions of 50% and higher, coupled with potential benefits in safety (from better braking and roll and stability control), may be possible with pneumatic (air blowing) devices ( Englar, 2001 [NPR] ). A complete package of aerodynamic improvements for a heavy-duty truck, including pneumatic blowing, might save about 15–20% of fuel for trucks operating primarily on uncongested highways, at a cost of about 5000 US$ in the near-term, with substantial cost reductions possible over time ( Vyas et al., 2002 [NPR, MoS] ).

The importance of aerodynamic forces at higher speeds implies that reduction of vehicle highway cruising speeds can save fuel and some nations have used speed limits as fuel conservation measures, e.g., the US during the period following the 1973 oil embargo. US tests on nine vehicles with model years from 1988 to 1997 demonstrated an average 17.1% fuel economy loss in driving at 70 mph compared to 55 mph ( ORNL, 2006 [NPR] ). Recent tests on six contemporary vehicles, including two hybrids, showed similar results – the average fuel economy loss was 26.5% in driving at 80 mph compared to 60 mph, and 27.2% in driving at 70 mph compared to 50 mph ( Duoba et al., 2005 [NPR] ).

Mobil Air Conditioning (MAC) systems

MAC systems contribute to GHG emissions in two ways by direct emissions from leakage of refrigerant and indirect emissions from fuel consumption. Since 1990 significant progress has been made in limiting refrigerant emissions due to the implementation of the Montreal Protocol. The rapid switch from CFC-12 (GWP 8100) to HFC-134a (GWP 1300 ) has led to the decrease in the CO₂-eq emissions from about 850 MtCO₂-eq in 1990 to 609 MtCO₂-eq in 2003, despite the continued growth of the MAC system fleet ( IPCC, 2005 [NPR] ).

Refrigerant emissions can be decreased by using new refrigerants with a much lower GWP, such as HFC-152a or CO₂, restricting refrigerant sales to certified service professionals and better servicing and disposal practices. Although the feasibility of CO₂ refrigerant has been demonstrated, a number of technical hurdles have still to be overcome.

Since the energy consumption for MAC is estimated to be 2.5–7.5% of total vehicle energy consumption, a number of solutions have to be developed in order to limit the energy consumption of MAC, such as improvements of the design of MAC systems, including the control system and airflow management.

5.3.1.2 Improving drive train efficiency

Advanced Direct Injection Gasoline / Diesel Engines and transmissions.

New engine and transmission technologies have entered the light-duty vehicle fleets of Europe, the USA and Japan, and could yield substantial reductions in carbon emissions if more widely used.

Direct injection diesel engines yielding about 35% greater fuel economy than conventional gasoline engines are being used in about half the light-duty vehicles being sold in European markets, but are little used in Japan and the USA (European taxes on diesel fuel generally are substantially lower than on gasoline, which boosts diesel share). Euro 4 emission standards were enforced in 2005, with Euro 5 (still undefined) to follow around 2009 – 2010 . These standards, plus Tier 2 standards in the USA, will challenge diesel NO_x controls, adding cost and possibly reducing fuel efficiency somewhat. Euro 4/Tier 2 compliant diesels for light-duty vehicles, obtaining 30% better fuel efficiency than conventional gasoline engines, may cost about 2000 –3000 US$ more than gasoline engines ( EEA, 2003 [NPR] ).

Improvements to gasoline engines include direct injection. Mercedes’ M271 turbocharged direct injection engine is estimated to attain 18% reduced fuel consumption, part of which is due to intake valve control and other engine technologies (SAE International, 2003a ); cylinder shutoff during low load conditions (Honda Odyssey V6, Chrysler Hemi, GM V8s) (SAE International, 2003a ) and improved valve timing and lift controls.

Transmissions are also being substantially improved. Mercedes, GM, Ford, Chrysler, Volkswagen and Audi are introducing advanced 6 and 7 speed automatics in their luxury vehicles, with strong estimated fuel economy improvements ranging from 4–8% over a 4-speed automatic for the Ford/GM 6-speed to a claimed 13% over a manual, plus faster acceleration, for the VW/Audi BorgWarner 6-speed (SAE International, 2003b ). If they follow the traditional path for such technology, these transmissions will eventually be rolled into the fleet. Also, continuously variable transmissions (CVTs), which previously had been limited to low power drive trains, are gradually rising in their power-handling capabilities and are moving into large vehicles.

The best diesel engines currently used in heavy-duty trucks are very efficient, achieving peak efficiencies in the 45–46% range USDOE,The best diesel engines currently used in heavy-duty trucks are very efficient, achieving peak efficiencies in the 45–46% range USDOE,). Although recent advances in engine and drive train technology for heavy-duty trucks have focused on emissions reductions, current research programmes in the US Department of Energy are aiming at 10–20% improvements in engine efficiency within ten years USDOE,). Although recent advances in engine and drive train technology for heavy-duty trucks have focused on emissions reductions, current research programmes in the US Department of Energy are aiming at 10–20% improvements in engine efficiency within ten years USDOE,), with further improvements of up to 25% foreseen if significant departures from the traditional diesel engine platform can be achieved.

Engines and drive trains can also be made more efficient by turning off the engine while idling and drawing energy from other sources. The potential for reducing idling emissions in heavy-duty trucks is significant. In the USA, a nationwide survey found that, on average, a long-haul truck consumed about 1,600 gallons, or 6,100 litres, per year from idling during driver rest periods. A variety of behavioural and technological practices could be pursued to save fuel. A technological fix is to switch to grid connections or use onboard auxiliary power units during idling ( Lutsey et al., 2004 [SRC] ).

Despite the continued tightening of emissions standards for both light-duty vehicles and freight trucks, there are remaining concerns about the gap between tested emissions and on-road emissions, particularly for diesel engines. Current EU emissions testing uses test cycles that are considerably gentler than seen in actual driving, allowing manufacturers to design drive trains so that they pass emissions tests but ‘achieve better fuel efficiency or other performance enhancement at the cost of higher emissions during operation on the road ( ECMT, 2006 [NPR, MoS] ).’ Other concerns involve excessive threshold limits demanded of onboard diagnostics systems, aftermarket mechanical changes (replacement of computer chips, disconnection of exhaust gas recirculation systems) and failure to maintain required fluid levels in Selective Catalytic Reduction systems ( ECMT, 2006 [NPR, MoS] ). Similar concerns in the USA led to the phase-in between 2000 and 2004 of a more aggressive driving cycle (the US06 cycle) to emission tests for LDVs; however, the emission limits tied to this cycle were not updated when new Tier 2 emission standards were promulgated, so concerns about onroad emissions, especially for diesels, will apply to the USA as well.

Hybrid drive trains

Hybrid-electric drive trains combine a fuel-driven power source, such as a conventional internal combustion engine (ICE) with an electric drive train – electric motor/generator and battery (or ultracapacitor) - in various combinations. ^[17] In current hybrids, the battery is recharged only by regenerative braking and engine charging, without external charging from the grid. ‘Plug-in hybrids,’ which would obtain part of their energy from the electric grid, can be an option but require a larger battery and perhaps a larger motor. Hybrids save energy by:

• Shutting the engine down when the vehicle is stopped (and possibly during braking or coasting);
• Recovering braking losses by using the electric motor to brake and using the electricity generated to recharge the battery;
• Using the motor to boost power during acceleration, allowing engine downsizing and improving average engine efficiency;
• Using the motor instead of the engine at low load (in some configurations), eliminating engine operation during its lowest efficiency mode;
• Allowing the use of a more efficient cycle than the standard Otto cycle (in some hybrids);
• Shifting power steering and other accessories to (more efficient) electric operation.

Since the 1998 introduction of the Toyota Prius hybrid in the Japanese market, hybrid electric drive train technology has advanced substantially, expanding its markets, developing in alternative forms that offer different combinations of costs and benefits and improving component technologies and system designs. Hybrids now range from simple belt-drive alternator-starter systems offering perhaps 7 or 8% fuel economy benefit under US driving conditions to ‘full hybrids’ such as the Prius offering perhaps 40–50% fuel economy benefits ^[18] (the Prius itself more than doubles the fuel economy average – on the US test – of the combined 2004 US model year compact and medium size classes, although some portion of this gain is due to additional efficiency measures). Also, hybrids may improve fuel efficiency by substantially more than this in congested urban driving conditions, so might be particularly useful for urban taxis and other vehicles making frequent stops. Hybrid sales have expanded rapidly: in the United States, sales were about 7,800 in 2000 and have risen rapidly, to 207,000 in 2005 ^[19] ; worldwide hybrid sales were about 541,000 in 2005 (IEA Hybrid Website, 2006 ).

Improvements made to the Prius since its introduction demonstrate how hybrid technology is developing. For example, the power density of Prius’s nickel metal hydride batteries has improved from 600 W/kg¹ in 1998 to 1250 W/kg¹ in 2004 - a 108% improvement. Similarly, the batteries’ specific energy has increased 37% during the same period ( EEA, 2003 [NPR] ). Higher voltage in the 2004 Prius allows higher motor power with reduced electrical losses and a new braking-by-wire system maximizes recapture of braking energy. The 1998 Prius compact sedan attained 42 mpg on the US CAFE cycle, with 0–60 mph acceleration time of 14.5 seconds; the 2004 version is larger (medium size) but attains 55 mpg and a 0–60 of 10.5 seconds. Prius-type hybrid systems will add about 4,000 US$ to the price of a medium sized sedan ( EEA, 2003 [NPR] ), but continued cost reduction and development efforts should gradually reduce costs.

Hybridization can yield benefits in addition to directly improving fuel efficiency, including (depending on the design) enhanced performance (with reduced fuel efficiency benefits in some designs), less expensive 4-wheel drive systems, provision of electric power for off-vehicle use (e.g., GM Silverado hybrid), and ease of introducing more efficient transmissions such as automated manuals (using the motor to reduce shift shock).

Hybrid drive trains’ strong benefits in congested stop-and-go travel mesh well with some heavier-duty applications, including urban buses and urban delivery vehicles. An initial generation of hybrid buses in New York City obtained about a 10% improvement in fuel economy as well as improved acceleration capacity and substantially reduced emissions ( Foyt, 2005 [NPR] ). More recently, a different design achieved a 45% fuel economy increase in NYC operation (not including summer, where the increase should be lower) ( Chandler et al., 2006 [NPR] ). Fedex has claimed a 57% fuel economy improvement for its E700 diesel hybrid delivery vehicles (Green Car Congress,). Fedex has claimed a 57% fuel economy improvement for its E700 diesel hybrid delivery vehicles (Green Car Congress,).

Hybrid applications extend to two and three-wheelers, as well, because these often operate in crowded urban areas in stop-and-go operation. Honda has developed a 50 cc hybrid scooter prototype that offers about a one-third reduction in fuel use and GHG emissions compared to similar 50 cc scooters Honda,Hybrid applications extend to two and three-wheelers, as well, because these often operate in crowded urban areas in stop-and-go operation. Honda has developed a 50 cc hybrid scooter prototype that offers about a one-third reduction in fuel use and GHG emissions compared to similar 50 cc scooters Honda,). However, sales of two and three-wheeled vehicles in most markets are extremely price sensitive, so the extent of any potential market for hybrid technology may be quite limited.

Plug-in hybrids, or PHEVs, are a merging of hybrid electric and battery electric. PHEVs get some of their energy from the electricity grid. Plug-in hybrid technology could be useful for both light-duty vehicles and for a variety of medium duty vehicles, including urban buses and delivery vehicles. Substantial market success of PHEV technology is, however, likely to depend strongly on further battery development, in particular on reducing battery cost and specific energy and increasing battery lifetimes.

PHEVs’ potential to reduce oil use is clear – they can use electricity to ‘fuel’ a substantial portion of miles driven. The US Electric Power Research Institute ( EPRI, 2001 [NPR] ) estimates that 30 km hybrids (those that have the capability to operate up to 30 km solely on electricity from the battery) can substitute electricity for gasoline for approximately 30–40% of miles driven in the USA. With larger batteries and motors, the vehicles could replace even more mileage. However, their potential to reduce GHG emissions more than that achieved by current hybrids depends on their sources of electricity. For regions that rely on relatively low-carbon electricity for off-peak power, e.g., natural gas combined cycle power, GHG reductions over the PHEV’s lifecycle will be substantial; in contrast, PHEVs in areas that rely on coal-fired power could have increased lifecycle carbon emissions. In the long-term, movement to a low-carbon electricity sector could allow PHEVs to play a major role in reducing transport sector GHG emissions.

5.3.1.3 Alternative fuels

Biofuels

The term biofuels describes fuel produced from biomass. A variety of techniques can be used to convert a variety of CO₂ neutral biomass feedstocks into a variety of fuels. These fuels include carbon-containing liquids such as ethanol, methanol, biodiesel, di-methyl esters (DME) and Fischer-Tropsch liquids, as well as carbon-free hydrogen. Figure 5.8 shows some main routes to produce biofuels: extraction of vegetable oils, fermentation of sugars to alcohol, gasification and chemical synthetic diesel, biodiesel and bio oil. In addition, there are more experimental processes, such as photobiological processes that produce hydrogen directly.

Figure 5.8: Overview of conversion routes from crops to biofuels

Source: Adapted from Hamelinck and Faaij, 2006 [ARC] .

Biofuels can be used either ‘pure’ or as a blend with other automotive fuels. There is a large interest in developing biofuel technologies, not only to reduce GHG emission but more so to decrease the enormous transport sector dependence on imported oil. There are two biofuels currently used in the world for transport purposes – ethanol and biodiesel.

Ethanol is currently made primarily by the fermentation of sugars produced by plants such as sugar cane, sugar beet and corn. Ethanol is used in large quantities in Brazil where it is made from sugar cane, in the USA where it is made from corn, but only in very small quantities elsewhere.

Ethanol is blended with gasoline at concentrations of 5–10% on a volume basis in North America and Europe. In Brazil ethanol is used either in its pure form replacing gasoline, or as a blend with gasoline at a concentration of 20–25%. The production of ethanol fuelled cars in Brazil achieved 96% market share in 1985, but sharply declining shortly thereafter to near zero. Ethanol vehicle sales declined because ethanol producers shifted to sugar production and consumers lost confidence in reliable ethanol supply. A 25% blend of ethanol has continued to be used. With the subsequent introduction of flexfuel cars (see Box 5.2 ), ethanol fuel sales have increased. However, the sugar cane experience in Brazil will be difficult to replicate elsewhere. Land is plentiful, the sugar industry is highly efficient, the crop residues (bagasse) are abundant and easily used for process energy, and a strong integrated R&D capability has been developed in cane growing and processing.

In various parts of Asia and Africa, biofuels are receiving increasing attention and there is some experience with ethanol-gasoline blending of up to 20%. Ethanol is being produced from sugar cane in Africa and from corn in small amounts in Asia. Biodiesel production is being considered from Jatropha (a drought resistant crop) that can be produced in most parts of Africa ( Yamba and Matsika, 2004 [NPR, ARC] ). It is estimated that with 10% ethanol-gasoline blending and 20% biodiesel-diesel blending in southern Africa, a reduction of 2.5 MtCO2 and 9.4 MtCO2 respectively per annum can be realized. Malaysian palm oil and US soybean oils are currently being used as biodiesel transport fuel in limited quantities and other oilseed crops are being considered elsewhere.

For the future, the conversion of ligno-cellulosic sources into biofuels is the most attractive biomass option. Ligno-cellulosic sources are grasses and woody material. These include crop residues, such as wheat and rice straw, and corn stalks and leaves, as well as dedicated energy crops. Cellulosic crops are attractive because they have much higher yields per hectare than sugar and starch crops, they may be grown in areas unsuitable for grains and other food/feed crops and thus do not compete with food, and the energy use is far less, resulting in much greater GHG reductions than with corn and most food crops ( IEA, 2006a [NPR] ).

A few small experimental cellulosic conversion plants were being built in the USA in 2006 to convert crop residues (e.g., wheat straw) into ethanol, but considerably more R&D investment is needed to make these processes commercial. These investments are beginning to be made. In 2006 BP announced it was committing 1 billion US$ to develop new biofuels, with special emphasis on bio-butanol, a liquid that can be easily blended with gasoline. Other large energy companies were also starting to invest substantial sums in biofuels R&D in 2006, along with the US Department of Energy, to increase plant yields, develop plants that are better matched with process conversion technologies and to improve the conversion processes. The energy companies in particular are seeking biofuels other than ethanol that would be more compatible with the existing petroleum distribution system. ^[22]

Biodiesel is less promising in terms of cost and production potential than cellulosic fuels but is receiving increasing attention. Bioesters are produced by a chemical reaction between vegetable or animal oil and alcohol, such as ethanol or methanol. Their properties are similar to those of diesel oil, allowing blending of bioesters with diesel or the use of 100% bioesters in diesel engines, and they are all called biodiesel. Blends of 20% biodiesel with 80% petroleum diesel (B20) can generally be used in unmodified diesel engines. ^[23]

Diesel fuel can also be produced through thermochemical hydrocracking of vegetable oil and animal fats. This technology has reached the demonstration stage. In Finland and Brazil ^[24] a commercial production project is under way. The advantage of the hydrocracked biodiesel is its stability and compatibility with conventional diesel ( Koyama et al., 2006 [NPR] ).

A large drawback of biodiesel fuels is the very high cost of feedstocks. If waste oils are used the cost can be competitive, but the quantity of waste oils is miniscule compared to transport energy consumption. If crops are used, the feedstock costs are generally far higher than for sugar, starch or cellulosic materials. These costs are unlikely to drop since they are the same highly developed crops used for foods and food processing. Indeed, if diverted to energy use, the oil feedstock costs are likely to increase still further, creating a direct conflict with food production. The least expensive oil feedstock at present is palm oil. Research is ongoing into new ways of producing oils. The promising feedstock seems to be algae, but cost and scale issues are still uncertain.

For 2030 IEA 2006a [NPR] ) reports mitigation potentials for bioethanol between 500– 1200 MtCO₂, with possibly up to 100–300 MtCO₂ of that for ligno-cellulosic ethanol (or some other bio-liquid). The long-term potential for ligno-cellulosic fuels beyond 2030 is even greater. For biodiesel, it reports mitigation potential between 100–300 MtCO₂.

The GHG reduction potential of biofuels, especially with cellulosic materials, is very large but uncertain. IEA estimated the total mitigation potential of biofuels in the transport sector in 2050 to range from 1800 to 2300 MtCO₂ at 25 US$/tCO₂-eq. based on scenarios with a respective replacement of 13 and 25% of transport energy demand by biofuels ( IEA, 2006a [NPR] ). The reduction uncertainty is huge because of uncertainties related to costs and GHG impacts.

Only in Brazil is biofuel competitive with oil at 50 US$ per barrel or less. All others cost more. As indicated in Figure 5.9 , biofuel production costs are expected to drop considerably, especially with cellulosic feedstocks. But even if the processing costs are reduced, the scale issue is problematic. These facilities have large economies of scale. However, there are large diseconomies of scale in feedstock production ( Sperling, 1985 [SRC] ). The cost of transporting bulky feedstock materials to a central point increases exponentially, and it is difficult assembling large amount of contiguous land to serve single large processing facilities.

Figure 5.9: Comparison of cost for various biofuels with those for gasoline and diesel

Source: IEA, 2006b [NPR] .

Another uncertainty is the well-to-wheel reduction in GHGs by these various biofuels. The calculations are very complex because of uncertainties in how to allocate GHG emissions across the various products likely to be produced in the bio-refinery facilities, how to handle the effects of alternative uses of land, and so on, and the large variations in how the crops are grown and harvested, as well as the uncertain efficiencies and design configurations of future process technologies and bio-engineering plant materials. Typical examples are shown in Figure 5.10 .

Figure 5.10: Reduction of well-to-wheels GHG emissions compared to conventionally fuelled vehicles

Note: bars indicate range of estimates.

Source: IEA, 2004c [NPR] ; EUCAR/CONCAWE/JRC, 2006 .

Ethanol from sugar cane, as produced in Brazil, provides significant reductions in GHG emissions compared to gasoline and diesel fuel on a ‘well-to-wheels’ basis. These large reductions result from the relatively energy efficient nature of sugar cane production, the use of bagasse (the cellulosic stalks and leaves) as process energy and the highly advanced state of Brazilian sugar farming and processing. Ongoing research over the years has improved crop yields, farming practices and process technologies. In some facilities the bagasse is being used to cogenerate electricity which is sold back to the electricity grid.

In contrast, the GHG benefits of ethanol made from corn are minor (Ribeiro & Yones-Ibrahim, 2001 [NotFound] ). Lifecycle estimates range from a net loss to gains of about 30%, relative to gasoline made from conventional oil. Farrell et al. 2006 [JoC] ) evaluates the many studies and concludes that on average the reductions are probably about 13% compared to gasoline from conventional oil. The corn-ethanol benefits are minimal because corn farming and processing are energy intensive.

Biofuels might play an important role in addressing GHG emissions in the transport sector, depending on their production pathway ( Figure 5.10 ). In the years to come, some biofuels may become economically competitive, as the result of increased biomass yields, developments of plants that are better suited to energy production, improved cellulosic conversion processes and even entirely new energy crops and conversion processes. In most cases, it will require entirely new businesses and industries. The example of ethanol in Brazil is a model. The question is the extent to which this model can be replicated elsewhere with other energy crops and production processes.

The biofuel potential is limited by:

• The amount of available agricultural land (and in case of competing uses for that land) for traditional and dedicated energy crops;
• The quantity of economically recoverable agricultural and silvicultural waste streams;
• The availability of proven and cost-effective conversion technology.

Another barrier to increasing the potential is that the production of biofuels on a massive scale may require deforestation and the release of soil carbon as mentioned in Chapter 8 .4. Another important point on biofuels is a view from the cost-effectiveness among the sectors. When comparing the use of biofuel in the transport sector with its use in power stations, the latter is more favourable from a cost-effectiveness point of view ( ECMT, 2007 [NPR, 2007] ).

Natural Gas (CNG / LNG / GTL)

Natural gas, which is mainly methane (CH₄), can be used directly in vehicles or converted into more compact fuels. It may be stored in compressed (CNG) or liquefied (LNG) form on the vehicle. Also, natural gas may be converted in large petrochemical plants into petroleum-like fuels (the process is known as GTL, or gas-to-liquid). The use of natural gas as a feedstock for hydrogen is described in the hydrogen section.

CNG and LNG combustion characteristics are appropriate for spark ignition engines. Their high octane rating, about 120, allows a higher compression ratio than is possible using gasoline, which can increase engine efficiency. This requires that the vehicle be dedicated to CNG or LNG, however. Many current vehicles using CNG are converted from gasoline vehicles or manufactured as bifuel vehicles, with two fuel tanks. Bifuel vehicles cannot take full advantage of CNG’s high octane ratio.

CNG has been popular in polluted cities because of its good emission characteristics. However, in modern vehicles with exhaust gas after-treatment devices, the non-CO₂ emissions from gasoline engines are similar to CNG, and consequently CNG loses its emission advantages in term of local pollutants; however it produces less CO₂. Important constraints on its use are the need for a separate refuelling infrastructure system and higher vehicle costs – because CNG is stored under high pressure in larger and heavier fuel tanks.

Gas-to-liquids (GTL) processes can produce a range of liquid transport fuels using Fischer-Tropsch or other conversion technologies. The main GTL fuel produced will be synthetic sulphur-free diesel fuel, although other fuels can also be produced. GTL processes may be a major source of liquid fuels if conventional oil production cannot keep up with growing demand, but the current processes are relatively inefficient: 61–65% (EUCAR/CONCAWE/JRC, 2006 ) and would lead to increased GHG emissions unless the CO₂ generated is sequestered.

DME can be made from natural gas, but it can also be produced by gasifying biomass, coal or even waste. It can be stored in liquid form at 5–10 bar pressure at normal temperature. This pressure is considerably lower than that required to store natural gas on board vehicles (200 bar). A major advantage of DME is its high cetane rating, which means that self-ignition will be easier. The high cetane rating makes DME suitable for use in efficient diesel engines.

DME is still at the experimental stage and it is still too early to say whether it will be commercially viable. During experiments, DME has been shown to produce lower emissions of hydrocarbons, nitric oxides and carbon monoxide than diesel and zero emissions of soot ( Kajitani et al., 2005 [NPR] ). There is no current developed distribution network for DME, although it has similarities to LPG and can use a similar distribution system. DME has a potential to reduce GHG emissions since it has a lower carbon intensity (15 tC/TJ) than petroleum products (18.9–20.2 tC/TJ) ( IPCC, 1996 [NPR] ).

Hydrogen / Fuel Cells

During the last decade, fuel cell vehicles (FCVs) have attracted growing attention and have made significant technological progress. Drivers for development of FCVs are global warming (FCVs fuelled by hydrogen have zero CO₂ emission and high efficiency), air quality (zero tailpipe emissions), and energy security (hydrogen will be produced from a wide range of sources), and the potential to provide new desirable customer attributes (low noise, new designs).

There are several types of FCVs; direct-drive and hybrid power train architectures fuelled by pure hydrogen, methanol and hydrocarbons (gasoline, naphtha). FCVs with liquid fuels have advantages in terms of fuel storage and infrastructure, but they need on-board fuel reformers (fuel processors), which leads to lower vehicle efficiency (30–50% loss), longer start-up time, slower response and higher cost. Because of these disadvantages and rapid progress on direct hydrogen systems, nearly all auto manufacturers are now focused on the pure hydrogen FCV. Significant technological progress has been made since TAR including: improved fuel cell durability, cold start (sub-freezing) operation, increased range of operation, and dramatically reduced costs (although FCV drive train costs remain at least an order of magnitude greater than internal combustion engine (ICE) drive train costs) ( Murakami and Uchibori, 2006 [NPR] ).

In addition, many demonstration projects have been initiated since TAR ^[25] . Since 2000, members of the California Fuel Cell Partnership have placed 87 light-duty FCVs and 5 FC buses in California, which have travelled over 590,000 km on California’s roads and highways. In 2002 – 2003, Japanese automakers began leasing FCVs in Japan and the USA, now totalling 17 vehicles. In 2004, US DOE started government/industry partnership ‘learning demonstrations’ for testing, demonstrating and validating hydrogen fuel cell vehicles and infrastructure and vehicle/infrastructure interfaces for complete system solutions. In Europe, there are several partnerships for FCV demonstration such as CUTE (Clean Urban Transport for Europe), CEP (Clean Energy Partnership) and ECTOS (Ecological City Transport System), using more than 30 buses and 20 passenger cars.

The recent US (NRC/NAE, 2004 ) and EU (JRC/IPTS, 2004 ) analyses conclude:

Although the potential of FCVs for reducing GHG emissions is very high there are currently many barriers to be overcome before that potential can be realized in a commercial market. These are:

• To develop durable, safe, and environmentally desirable fuel cell systems and hydrogen storage systems and reduce the cost of fuel cell and storage components to be competitive with today’s ICEs;
• To develop the infrastructure to provide hydrogen for the light-duty vehicle user;
• To sharply reduce the costs of hydrogen production from renewable energy sources over a time frame of decades. Or to capture and store (‘sequester’) the carbon dioxide byproduct of hydrogen production from fossil fuels.

Public acceptance must also be secured in order to create demand for this technology. The IEA echoes these points while also noting that deployment of large-scale hydrogen infrastructure at this point would be premature, as some of the key technical issues that are still being worked on, such as fuel cell operating conditions and hydrogen on-board storage options, may have a considerable impact on the choice of hydrogen production, distribution and refuelling ( IEA, 2005 [NPR] ).

The GHG impact of FCVs depends on the hydrogen production path and the technical efficiency achieved by vehicles and H₂ production technology. At the present technology level with FCV tank-to-wheel efficiency of about 50% and where hydrogen can be produced from natural gas at 60% efficiency, well-to-wheel (WTW) CO₂ emissions can be reduced by 50–60% compared to current conventional gasoline vehicles. In the future, those efficiencies will increase and the potential of WTW CO₂ reduction can be increased to nearly 70%. If hydrogen is derived from water by electrolysis using electricity produced using renewable energy such as solar and wind, or nuclear energy, the entire system from fuel production to end-use in the vehicle has the potential to be a truly ‘zero emissions’. The same is almost true for hydrogen derived from fossil sources where as much as 90% of the CO₂ produced during hydrogen manufacture is captured and stored (see Figure 5.11 ).

Figure 5.11: Well-to-wheel CO₂ emission for major pathways of hydrogen with some estimates of hydrogen production cost (numbers in parentheses)

Source: Toyota/Mizuho, 2004; NRC/NAE, 2004 .

FCV costs are expected to be much higher than conventional ICE vehicles, at least in the years immediately following their introduction and H₂ costs may exceed gasoline costs. Costs for both the vehicles and fuel will almost certainly fall over time with larger-scale production and the effects of learning, but the long-term costs are highly uncertain. Figure 5.11 shows both well-to-wheels emissions estimates for several FCV pathways and their competing conventional pathways, as well as cost estimates for some of the hydrogen pathways.

Although fuel cells have been the primary focus of research on potential hydrogen use in the transport sector, some automakers envision hydrogen ICEs as a useful bridge technology for introducing hydrogen into the sector and have built prototype vehicles using hydrogen. Mazda has started to lease bi-fuel (hydrogen or gasoline) vehicles using rotary engines and BMW has also converted a 7-series sedan to bi-fuel operation using liquefied hydrogen ( Kiesgen et al., 2006 [NPR] ) and is going to lease them in 2007 . Available research implies that a direct injected turbocharged hydrogen engine could potentially achieve efficiency greater than a di diesel ( Wimmer et al., 2005 [NPR] ), although research and development challenges remain, including advanced sealing technology to insure against leakage with high pressure injection.

Electric vehicles

Fuel cell and hybrid vehicles gain their energy from chemical fuels, converting them into electricity onboard. Pure electric vehicles operating today are either powered from off-board electricity delivered through a conductive contact – usually buses with overhead wires or trains with electrified ‘third rails’ – or by electricity acquired from the grid and stored on-board in batteries. Future all-electric vehicles might use inductive charging to acquire electricity, or use ultracapacitors or flywheels in combination with batteries to store electricity on board.

The electric vehicles are driven by electric motors with high efficiencies of more than 90%, but their short driving range and short battery life have limited the market penetration. Even a limited driving range of 300 km requires a large volume of batteries weighing more than 400 kg ( JHFC, 2006 [NPR] ). Although the potential of CO₂ reduction strongly depends on the power mix, well-to-wheels CO₂ emission can be reduced by more than 50% compared to conventional gasoline-ICE ( JHFC, 2006 [NPR] ).

Vehicle electrification requires a more powerful, sophisticated and reliable energy-storage component than lead-acid batteries. These storage components will be used to start the car and also operate powerful by-wire control systems, store regenerative braking energy and to operate the powerful motor drives needed for hybrid or electric vehicles. Nickel metal hydride (NiMH) batteries currently dominate the power-assist hybrid market and Li ion batteries dominate the portable battery business. Both are being aggressively developed for broader automotive applications. The energy density has been increased to 170 Wh kg^–1 and 500 Wh L^–1 for small-size commercial Li ion batteries (Sanyo, 2005 ) and 130 Wh kg^–1 and 310 Wh L^–1 for large-size EV batteries (Yuasa, 2000 ). While NiMH has been able to maintain hybrid vehicle high-volume business, Li ion batteries are starting to capture niche market applications (e.g., the idle-stop model of Toyota’s Vitz). The major hurdle left for Li ion batteries is their high cost.

Ultracapacitors offer long life and high power but low energy density and high current cost. Prospects for cost reduction and energy enhancement and the possibility of coupling the capacitor with the battery are attracting the attention of energy storage developers and automotive power technologists alike. The energy density of ultracapacitors has increased to 15–20 Wh kg^–1 (Power System, 2005 [NPR] ), compared with 40–60 Wh kg^–1 for Ni-MH batteries. The cost of these advanced capacitors is in the range of several 10s of dollars/Wh, about one order of magnitude higher than Li batteries.

5.3.1.4 Well-to-wheels analysis of technical mitigation options

Life cycle analysis (LCA) is the most systematic and comprehensive method for the assessment of the environmental impacts of transport technologies. However, non-availability, uncertainty or variability of data limit its application. One key difficulty is deciding where to draw the boundary for the analysis; another is treating the byproducts of fuel production systems and their GHG emission credits. Also in some cases, LCA data varies strongly across regions

For automobiles, the life-cycle chain can be divided into the fuel cycle (extraction of crude oil, fuel processing, fuel transport and fuel use during operation of vehicle) and vehicle cycle (material production, vehicle manufacturing and disposal at the end-of-life). For a typical internal combustion engine (ICE) vehicle, 70–90% of energy consumption and GHG emissions take place during the fuel cycle, depending on vehicle efficiency, driving mode and lifetime driving distance (Toyota, 2004 ).

Recent studies of the Well-to-wheels CO₂ emissions of conventional and alternative fuels and vehicle propulsion concepts include a GM/ANL ( 2005 ) analysis for North America, EU-CAR/CONCAWE/JRC ( 2006 ) for Europe and Toyota/Mizuho ( 2004 ) for Japan. Some results are shown in Fig. 5.12. Some of the differences, as apparent from Figure 5.12 for ICE-gasoline and ICE-D (diesel) reflect difference in the oil producing regions and regional differences in gasoline and diesel fuel requirements and processing equipment in refineries.

The Well-to-wheel CO₂ emissions shown in Fig. 5.12 are for three groups of vehicle/fuel combinations – ICE/fossil fuel, ICE/biofuel and FCV. The full well-to-wheels CO₂ emissions depend on not only the drive train efficiency (TTW: tank-to-wheel) but also the emissions during the fuel processing (WTT: well-to-tank). ICE-CNG (compressed natural gas) has 15–25% lower emissions than ICE-G (Gasoline) because natural gas is a lower-carbon fuel and ICE-D (Diesel) has 16–24% lower emissions due to the high efficiency of the diesel engine. The results for hybrids vary among the analyses due to different assumptions of vehicle efficiency and different driving cycles. Although Toyota’s analysis is based on Prius, and using Japanese 10–15 driving cycle, the potential for CO₂ reduction is 20–30% in general.

Table 5.3 summarizes the results and provides an overview of implementation barriers. The lifecycle emissions of ICE vehicles using biofuels and fuel cell vehicles are extremely dependent on the fuel pathways. For ICE-Biofuel, the CO₂ reduction potential is very large (30–90%), though world potential is limited by high production costs for several biomass pathways and land availability. The GHG reduction potential for the natural gas-sourced hydrogen FCV is moderate, but lifecycle emissions can be dramatically reduced by using CCS (carbon capture and storage) technology during H₂ production (FCEV-H2ccs in Table 5.3 ). Using renewable energy such as C-neutral biomass as a feedstock or clean electricity as an energy source (FCEV-RE-H₂) also will yield very low emissions.

Table 5.3: Reduction of Well-to-wheels GHG emissions for various drive train/fuel combinations

5.3.1.5 Road transport: mode shifts

Personal motor vehicles consume much more energy and emit far more GHGs per passenger-km than other surface passenger modes. And the number of cars (and light trucks) continues to increase virtually everywhere in the world. Growth in GHG emissions can be reduced by restraining the growth in personal vehicle ownership. Such a strategy can, however, only be successful if high levels of mobility and accessibility can be provided by alternative means.

In general, collective modes of transport use less energy and generate less GHGs than private cars. Walking and biking emit even less. There is important worldwide mitigation potential if public and non-motorised transport trip share loss is reversed. The challenge is to improve public transport systems in order to preserve or augment the market share of low-emitting modes. If public transport gets more passengers, it is possible to increase the frequency of departures, which in turn may attract new passengers ( (Akerman and Hojer, 2006 ) ).

The USA is somewhat of an anomaly, though. In the USA, passenger travel by cars generates about the same GHG emissions as bus and air travel on a passenger-km basis (ORNL Transportation Energy Databook; ORNL, 2006 [NPR] ). That is mostly because buses have low load factors in the USA. Thus, in the USA, a bus-based strategy or policy will not necessarily lower GHG emissions. Shifting passengers to bus is not simply a matter of filling empty seats. To attract more passengers, it is necessary to enhance transit service. That means more buses operating more frequently – which means more GHG emissions. It is even worse than that, because transit service is already offered where ridership ^[26] demand is greatest. Adding more service means targeting less dense corridors or adding more service on an existing route. There are good reasons to promote transit use in the USA, but energy use and GHGs are not among them.

Virtually everywhere else in the world, though, transit is used more intensively and therefore has a GHG advantage relative to cars. Table 5.4 shows the broad average GHG emissions from different vehicles and transport modes in a developing country context. GHG emissions per passenger-km are lowest for transit vehicles and two-wheelers. It also highlights the fact that combining alternative fuels with public transport modes can reduce emissions even further.

It is difficult to generalize, though, because of substantial differences across nations and regions. The types of buses, occupancy factors, and even topography and weather can affect emissions. For example, buses in India and China tend to be more fuel-efficient than those in the industrialized world, primarily because they have considerably smaller engines and lack air conditioning ( Sperling and Salon, 2002 [NPR, SRC] ).

Table 5.4: GHG Emissions from vehicles and transport modes in developing countries

Public transport

In addition to reducing transport emissions, public transport is considered favourably from a socially sustainable point of view because it gives higher mobility to people who do not have access to car. It is also attractive from an economically sustainable perspective since public transport provides more capacity at less marginal cost. It is less expensive to provide additional capacity by expanding bus service than building new roads or bridges. The expansion of public transport in the form of large capacity buses, light rail transit and metro or suburban rail can be feasible mitigation options for the transport sector.

The development of new rail services can be an effective measure for diverting car users to carbon-efficient mode while providing existing public transport users with upgraded service. However, a major hurdle is higher capital and possibly operating cost of the project. Rail is attractive and effective at generating high ridership in very dense cities. During the 1990 s, less capital-intensive public transport projects such as light rail transit (LRT) were planned and constructed in Europe, North America and Japan. The LRT systems were successful in some regions, including a number of French cities where land use and transport planning is often well integrated ( Hylen and Pharoah, 2002 [NPR] ), but less so in other cities especially in the USA ( (Richmond, 2001; ) ( Mackett and Edwards, 1998 ) ), where more attention has been paid to this recently.

Around the world, the concept of bus rapid transit (BRT) is gaining much attention as a substitute for LRT and as an enhancement of conventional bus service. BRT is not new. Plans and studies for various BRT type alternatives have been prepared since the 1930 s and a major BRT system was installed in Curitiba, Brazil in the 1970 s ( (Levinson et al., 2002 ) ). But only since about 2000 has the successful Brazilian experience gained serious attention from cities elsewhere.

BRT is ‘a mass transit system using exclusive right of way lanes that mimic the rapidity and performance of metro systems, but utilizes bus technology rather than rail vehicle technology’ ( Wright, 2004 [NPR] ). BRT systems can be seen as enhanced bus service and an intermediate mode between conventional bus service and heavy rail systems. BRT includes features such as exclusive right of way lanes, rapid boarding and alighting, free transfers between routes and preboard fare collection and fare verification, as well as enclosed stations that are safe and comfortable, clear route maps, signage and real-time information displays, modal integration at stations and terminals, clean vehicle technologies and excellence in marketing and customer service. To be most effective, BRT systems (like other transport initiatives) should be part of a comprehensive strategy that includes increasing vehicle and fuel taxes, strict land-use controls, limits and higher fees on parking, and integrating transit systems into a broader package of mobility for all types of travellers ( IEA, 2002b [NPR, MoS] ).

Most BRT systems today are being delivered in the range of 1–15 million US$/km, depending upon the capacity requirements and complexity of the project. By contrast, elevated rail systems and underground metro systems can cost from 50 million US$ to over 200 million US$/km ( Wright, 2004 [NPR] ). BRT systems now operate in several cities throughout North America, Europe, Latin America, Australia, New Zealand and Asia. The largest and most successful systems to date are in Latin America in Bogotá, Curitiba and Mexico City ( Karekezi et al., 2003 [NPR] ).

Analysing the Bogotá Clean Development Mechanism project gives an insight into the cost and potential of implementing BRT in large cities. The CDM project shows the potential of moving about 20% of the city population per day on the BRT that mainly constitutes putting up dedicated bus lanes (130 km), articulated buses ( 1200 ) and 500 other large buses operating on feeder routes. The project is supported by an integrated fare system, centralized coordinated fleet control and improved bus management ^[27] . Using the investment costs, an assumed operation and maintenance of 20–50% ^[28] of investment costs per year, fuel costs of 40 to 60 US$ per barrel in 2030 and a discount rate of 4%, a BRT lifespan of 30 years, the cost of implementing BRT in the city of Bogotá was estimated to range from 7.6 US$/tCO₂ to 15.84 US$/tCO₂ depending on the price of fuel and operation and maintenance ( Table 5.5 ). Comparing with results of Winkelman 2006 [NPR] ), BRT cost estimates ranged from 14-66 US$/tCO₂ depending on the BRT package involved ( Table 5.6 ). The potential for CO₂ reduction for the city of Bogotá was determined to average 247,000 tCO₂ per annum or 7.4 million tCO₂ over a 30 year lifespan of the project.

Table 5.5: Cost and potential estimated for BRT in Bogota

Table 5.6: CO₂ reduction potential and cost per tCO₂ reduced using public transit policies in typical Latin American cities

Non-motorized transport (NMT)

The prospect for the reduction in CO₂ emissions by switching from cars to non-motorized transport (NMT) such as walking and cycling is dependent on local conditions. In the Netherlands, where 47% of trips are made by NMT, the NMT plays a substantial role up to distances of 7.5 km and walking up to 2.5 km ( Rietveld, 2001 [NPR] ). As more than 30% of trips made in cars in Europe cover distances of less than 3 km and 50% are less than 5 km ( EC, 1999 [NPR] ), NMT can possibly reduce car use in terms of trips and, to a lesser extent, in terms of kilometres. While the trend has been away from NMT, there is considerable potential to revive interest in NMT. In the Netherlands, with strong policies and cultural commitment, the modal share of bicycle and walking for accessing trains from home is about 35 to 40% and 25% respectively ( Rietveld, 2001 [NPR] ).

Walking and cycling are highly sensitive to the local built environment ( ECMT, 2004a [NPR] ;( Lee and Mouden, 2006 ) ). In Denmark, where the modal share of cycling is 18%, urban planners seek to enhance walking and cycling by shortening journey distances and providing better cycling infrastructure ( (Dill and Carr 2003, ) Page, 2005 [NPR] ). In the UK where over 60% of people live within a 15 minute bicycle ride of a station, NMT could be increased by offering convenient, secure bicycle parking at stations and improved bicycle carriage on trains ( ECMT, 2004a [NPR] ).

Safety is an important concern. NMT users have a much higher risk per trip of being involved in an accident than those using cars, especially in developing countries where most NMT users cannot afford to own a car ( (Mohan and Tiwari, 1999 ) ). Safety can be improved through traffic engineering and campaigns to educate drivers. An important co-benefit of NMT, gaining increasing attention in many countries, is public health (National Academies studies in the USA; Pucher, 2004 [NPR] ).

In Bogotá, in 1998, 70% of the private car trips were under 3 km. This percentage is lower today thanks to the bike and pedestrian facilities. The design of streets was so hostile to bicycle travel that by 1998 bicycle trips accounted for less than 1% of total trips. After some 250 km of new bicycle facilities were constructed by 2001 ridership had increased to 4% of total trips. In most of Africa and in much of southern Asia, bicyclists and other non-motorised and animal traction vehicles are generally tolerated on the roadways by authorities. Non-motorised goods transport is often important for intermodal goods transport. A special form of rickshaw is used in Bangladesh, the bicycle van, which has basically the same design as a rickshaw ( Hook, 2003 [NPR] ).

Mitigation potential of modal shifts for passenger transport

Rapid motorization in the developing world is beginning to have a large effect on global GHG emissions. But motorization can evolve in quite different ways at very different rates. The amount of GHG emissions can be considerably reduced by offering strong public transport, integrating transit with efficient land use, enhancing walking and cycling, encouraging minicars and electric two-wheelers and providing incentives for efficient vehicles and low-GHG fuels. Few studies have analyzed the potential effect of multiple strategies in developing nations, partly because of a severe lack of reliable data and the very large differences in vehicle mix and travel patterns among varying areas.

( Wright and Fulton 2005 ) ) estimated that a 5% increase in BRT mode share against a 1% mode share decrease of private automobiles, taxis and walking, plus a 2% share decrease of mini-buses can reduce CO₂ emissions by 4% at an estimated cost of 66 US$/tCO₂ in typical Latin American cities. A 5% or 4% increase in walking or cycling mode share in the same scenario analysis can also reduce CO₂ emissions by 7% or 4% at an estimated cost of 17 or 15 US$/tCO₂, respectively (Table5.6). Although the assumptions of a single infrastructure unit cost and its constant impact on modal share in the analysis might be too simple, even shifting relatively small percentages of mode share to public transport or NMT can be worthwhile, because of a 1% reduction in mode share of private automobiles represents over 1 MtCO₂ through the 20-year project period.

Figure 5.13 shows the GHG transport emission results, normalized to year 2000 emissions, of four scenario analyses of developing nations and cities ( Sperling and Salon, 2002 [NPR, SRC] ). For three of the four cases, the ‘high’ scenarios are ‘business-as-usual’ scenarios assuming extrapolation of observable and emerging trends with an essentially passive government presence in transport policy. The exception is Shanghai, which is growing and changing so rapidly that ‘business-as-usual’ has little meaning. In this case the high scenario assumes both rapid motorization and rapid population increases, with the execution of planned investments in highway infrastructure while at the same time efforts to shift to public transport falter ( Zhou and Sperling, 2001 [NPR, SRC] ).

Figure 5.13: Projections for transport GHG emissions in 2020 for some cities of developing countries

Notes: Components of the Low 2020 scenario:

Delhi ( Bose and Sperling, 2001 [NPR, SRC] ): Completion of planned busways and rail transit, land-use planning for high density development around railway stations, network of dedicated bus lanes, promotion of bicycle use, including purchase subsidies and special lanes, promotion of car sharing, major push for more natural gas use in vehicles, economic re-straints on personal vehicles.

Shanghai ( Zhou and Sperling, 2001 [NPR, SRC] ): Emphasis on rapid rail system growth, high density development at railway sta-tions, bicycle promotion with new bike lanes and parking at transit stations, auto industry focus on minicars and farm cars rather than larger vehicles, incentives for use of high tech in minicars – electric, hybrid, fuel cell drive trains, promotion of car sharing.

Chile ( O’Ryan et al., 2002 [NPR, SRC] ): Overall focus on stronger use of market-based policy to insure that vehicle users pay the full costs of driving, internalizing costs of pollution and congestion, parking surcharges and restrictions, vehicle fees, and road usage fees, improvements in bus and rail systems, encouragement of minicars, with lenient usage and parking rules and strong commitment to alternative fuels, especially natural gas. By 2020, all taxis and 10% of other light and medium vehicles will use natural gas; all new buses will use hydrogen, improvements in bus and rail sys-tems.

South Africa ( Prozzi et al., 2002 [NPR, SRC] ): Land-use policies towards more efficient growth patterns, strong push to improve public transport, including use of busways in dense corridors, provision of new and better buses, strong government oversight of the minibus jitney industry, incentives to moderate private car use, coal-based synfuels shifts to imported natural gas as a feedstock

Source: Sperling and Salon, 2002 [NPR, SRC] .

5.3.1.6 Improving driving practices (eco-driving)

Fuel consumption of vehicles can be reduced through changes in driving practices. Fuel-efficient driving practices, with conventional combustion vehicles, include smoother deceleration and acceleration, keeping engine revolutions low, shutting off the engine when idling, reducing maximum speeds and maintaining proper tyre pressure ( IEA, 2001 [NPR] ). Results from studies conducted in Europe and the USA suggested possible improvement of 5–20% in fuel economy from eco-driving training. The mitigation costs of CO₂ by eco-driving training were mostly estimated to be negative (ECMT/IEA, 2005 ).

Eco-driving training can be attained with formal training programmes or on-board technology aids. It applies to drivers of all types of vehicles, from minicars to heavy-duty trucks. The major challenge is how to motivate drivers to participate in the programme, and how to make drivers maintain an efficient driving style long after participating ( IEA, 2001 [NPR] ). In the Netherlands, eco-driving training is provided as part of driving school curricula (ECMT/IEA, 2005 ).

5.4.2.1 Light-duty Vehicles

The following estimate of the overall GHG emissions reduction potential and costs for improving the efficiency of the world’s light-duty vehicle fleet (thus reducing carbon emissions), is based on the IEA Reference Case, as documented in a spreadsheet model developed by the IEA for the Mobility 2030 project ( WBSCD, 2004b [NPR, MoS] ). The cost estimates for total mitigation potential are provided in terms of ‘societal’ costs of reductions in GHG emissions, measured in US$/tonne of carbon (tC) or carbon dioxide (CO₂); the costs are the net of higher vehicle costs minus discounted lifetime fuel savings. Fuel savings benefits are measured in terms of the untaxed cost of the fuels at the retail level, and future savings are discounted at a low societal rate of 4% per year. These costs are not the same as those that would be faced by consumers, who would face the full taxed costs of fuel, would almost certainly use a higher discount rate, and might value only a few years of fuel savings. Also, they do not include the consumer costs of forgoing further increases in vehicle performance and weight. Over the past few decades, increasing acceleration performance and vehicle weight have stifled increases in fuel economy for light-duty vehicles and these trends must be stopped if substantial progress is to be made in fleet efficiency. Because consumers value factors such as vehicle performance, stopping these trends will have a perceived cost – but there is little information about its magnitude.

The potential improvements in light-duty fuel economy assumed in the analysis, and the costs of these improvements, are based on the scenarios in the MIT study summarised in Box 5.5 . The efficiency improvements as mentioned in this study are discounted somewhat to take into account the period in which the full benefits can be achieved. Further, fleet penetration of the technology advances are assumed to be delayed by 5 years in developing nations; however, because developing nation fleets are growing rapidly, higher efficiency vehicles, once introduced, may become a large fraction of the total fleet in these nations within a relatively short time. The technology assumptions for two ‘efficiency scenarios’ are as follows ( Table 5.9 a).

Table 5.9a: Fuel economy and cost assumptions for cost and potentials analysis

The high efficiency and medium efficiency scenarios achieve the following improvements in efficiency for the new light-duty vehicle fleet ( Table 5.9 b):

Table 5.9b: Efficiency improvements new light-duty vehicle fleet

Table 5.10 shows the light-duty vehicle fuel consumption and (vehicle only) CO₂ emissions for the Reference scenario and the High and medium efficiency scenarios. In the Reference case, LDV fuel consumption increases by nearly 60% by 2030; the High Efficiency Case cuts this increase to 26% and the Medium efficiency scenario cuts it to 42%. For the OECD nations, the Reference Case projects only a 22% increase by 2030, primarily because of moderate growth in travel demand, with the High efficiency scenario actually reducing fuel consumption in this group of nations by 9% and the Medium efficiency scenario reducing growth to only 6%. This regional decrease (or modest increase) in fuel use is overwhelmed by the rapid growth in the world’s total fleet size and overall travel demand and the slower uptake of efficiency technologies in the developing nations. Because no change in the use of biofuels was assumed in this analysis, the CO₂ emissions in the scenarios essentially track the energy consumption paths discussed above. Figure 5.16 shows the GHG emissions path for the three scenarios, resulting in a mitigation potential of about 800 (High) and 400 (Medium) MtCO₂ in 2030 .

Table 5.10: Regional and worldwide Light-duty vehicle CO₂ emissions (vehicle only) and fuel consumption, efficiency and reference cases

Figure 5.16: Light-duty vehicle CO₂ emissions for three scenarios

Table 5.11 shows the cost of the reductions in GHG emissions in US$/tCO₂ for those reductions obtained by the 2030 new vehicle fleet over its lifetime, assuming oil prices of 30 US$, 40 US$, 50 US$ and 60 US$/bbl over the vehicles’ lifetime. ^[36] Note that the costs in Table 5.11 do not apply to the carbon reductions achieved in that year by the entire LDV fleet (from Table 5.10 ), because those reductions are associated with successive waves of high efficiency vehicles entering the fleet during the approximately 15 year period before (and including) 2030 .

Table 5.11: Cost of CO₂ reduction in new 2030 LDVs

The Table 5.11 results show that the ‘social cost of carbon reduction’ for light-duty vehicles varies dramatically across regions and with fuel prices (since the cost is the net of technology costs minus the value of fuel savings). The results are also quite different for the High and Medium efficiency scenarios, primarily because the estimated technology costs begin to rise more steeply at higher efficiency levels, raising the average cost/tonne of CO₂ in the High efficiency scenario. For the High efficiency scenario, CO₂ reduction costs are very high for the OECD countries aside from North America, even at 60 US$/bbl oil prices, reflecting the ambitious (and expensive) increases in that scenario, the relatively high efficiencies of those regions’ fleets in the Reference Case, and the relatively low km/vehicle/year driven outside North America; on the other hand, the costs of the moderate increases in the Medium efficiency scenario are low to negative for all regions, reflecting the availability of moderate cost technologies capable of raising average vehicle efficiencies up to 30–40% or so.

The values in Table 5.11 are sensitive to several important assumptions:

• Technology costs: the costs assumed here appear to be considerably higher than those assumed in WEO 2006 (IEA, 2006a).
• Discount rates: the analysis assumes a low social discount rate of 4% in keeping with the purpose of the analysis. As noted, vehicle purchasers would undoubtedly use higher rates and would value fuel savings at retail fuel prices rather than the untaxed values used here; they might also only value a few years of fuel savings rather than the lifetime savings assumed here. WEO 2006 on the other hand, used a zero discount rate, substantially reducing the net cost of carbon reduction.
• Vehicle km travelled (vkt): this analysis used the IEA/WBCSD spreadsheet’s assumption of constant vkt over time and applied these values to new cars. Actual driving patterns will depend on the balance of increasing road infrastructure and rapidly increasing fleet size in developing nations. Unless infrastructure keeps pace with growing fleet size, which will be difficult, the assumption of constant vkt/vehicle may prove accurate or even optimistic.
• Efficiency gains assumed in the Reference scenario: the Reference scenario assumed significant gains in most areas (aside from North America), which makes the Efficiency scenarios more expensive.

Table 5.12 shows the economic potential for reducing CO₂ emissions in the 2030 fleet of new LDVs as a function of world oil price. ^[37] The values show that much of the economic potential is available at a net savings, ‘if consumer preference for power and other efficiency-robbing vehicle attributes is ignored’. Even at 30 US$/bbl oil prices, over half of the total (<100 US$/tCO₂) potential is available at a net savings over the vehicle lifetime; at 40 US$/bbl, over 90% of the 718 Mt total potential is available at a net savings.

Table 5.12: Economic potential of LDV mitigation technologies as a function of world oil price, for new vehicles in 2030

The regional detail, not shown in Table 5.12 , is illuminating. In the High Efficiency scenario, of 793 Mt of total potential, 445 Mt are in OECD North America and are available at a net savings at 40 US$/bbl oil (and at less than 20 US$/tCO₂ at 30 US$/bbl oil). The next highest regional potential is in OECD Europe at 104 Mt, but this potential is more expensive: at 30 US$/bbl oil. Only 56 Mt is available below 100 US$/tCO₂, and becomes available at below 100 US$/tCO₂ only at 60 US$/bbl oil. China has the next highest total emissions ( 2030 Reference case emissions of 303 Mt) but only a moderate potential of 36 Mt. This potential is fully available at a net savings only if oil is 50 US$/bbl or higher – perhaps not surprising because China has ambitious fuel economy standards embedded in the Reference Case and has relatively low driving rates, which make further improvements more difficult and expensive.

5.4.2.2 Aircraft

QinetiQ (UK) ^[38] analysed the fuel consumption and CO₂ trends for a simple global aviation growth scenario to provide an indicative view on the extent that technology and other developments might mitigate aviation emissions. The ICAO traffic forecast (ICAO/FESG, 2003 ) defined traffic growth to 2030 from which a future fleet composition was developed, using a range of current and future aircraft types where their introduction could be assumed, as well as representative aircraft types based on seat capacity. Fuel burn and emissions were calculated using known emissions performance and projections for future aircraft where necessary.

The analysis assumed a range of technology options as follows:

• Case 1 assumed no technology change from 2002 to 2030; using the extrapolated traffic forecast from ICAO FESG – this case shows only the effects of traffic growth on emissions.
• Case 2 – as Case 1, but assumes all new aircraft deliveries after 2005 would be ‘best available technology at a 2005 (BAT)’ performance standard, and with specific new aircraft (A380, A350, B787) delivered from 2008.
• Case 3 – as Case 1, but with assumed annual fleet fuel efficiency improvements as per ‘Greene’ and DTI (IPCC 1999, Chapter 9, Table 9.15). This assumes a fleet efficiency improvement trend of 1.3% per year to 2010, assumed then to decline to 1.0% per year to 2020 and 0.5% per year thereafter. This is the reference case.
• Case 4 – as Case 3, plus the assumption that a 50 US$/tCO₂ cost will produce a further 0.5% fuel efficiency improvement per annum from 2005, as suggested by the cost-potential estimates of Wit et al., (2002), that assume technologies such as winglets, fuselage skin treatments (riblets) and further weight reductions and engine developments will be introduced by airlines.
• Case 5 – as Case 3, plus the assumption of 100 US$/tCO₂ cost, producing a 1.0% fuel efficiency improvement per annum from 2005 (Wit et al., 2002), again influencing the introduction of additional technologies as above.

The results of this analysis are summarised in Table 5.13 .

Table 5.13: Summaries of CO₂ mitigation potential analysis in aviation

Case 2 is a simple representation of planned industry developments and shows their effect to 2030, ignoring further technology developments. This is an artificial case, as on-going efficiency improvements would occur as a matter of course, but it shows that these planned fleet developments alone might save 14% of the CO₂ that the ‘no technology change’ of Case 1 would have produced. Case 3 should be regarded as the ‘base case’ from which benefits are measured, as this case reflects an agreed fuel efficiency trend assumed for some of the calculations produced in the IPCC Special Report ( 1999 ). This results in a further 11% reduction in CO₂ by 2030 compared with Case 2. Cases 4 and 5 assume that a carbon cost will drive additional technology developments from 2005 – no additional demand effect has been assumed. These show further CO₂ reduction of 11.8% and 22.2% compared with ‘base case’ 3 over the same period from technologies that are assumed to be more attractive than hitherto. However, even the most ambitious scenario suggests that CO₂ production will increase by almost 100% from the base year. The cost potentials for Cases 4 and 5 are based on one study and further studies may refine these results. There is limited literature in the public domain on costs of mitigation technologies. The effects of more advanced technology developments, such as the blended wing body, are not modelled here, as these developments are assumed to take place after 2030 .

The analysis suggests that aviation emissions will continue to grow as a result of continued demand for civil aviation. Assuming the historical fuel efficiency trend produced by industry developments will continue (albeit at a declining level), carbon emissions will also grow, but at a lower rate than traffic. Carbon pricing could effect further emissions reductions if the aviation industry introduces further technology measures in response.

5.4.2.3 Biofuels

IEA has projected the potential worldwide increased use of biofuels in the transport sector assuming successful technology development and policy measures reducing barriers to biomass deployment and providing economic incentives.

IEA’s World Energy Outlook 2006 ( IEA, 2006b [NPR] ) develops an Alternative policy scenario that adds 55 Mtoe biofuels above baseline levels of 92 Mtoe by 2030, which increases the biofuels share of total transport fuel demand from 3 to 5%. In this scenario, all of the biofuels are produced by conventional technology, that is ethanol from starch and sugar crops and biodiesel from oil crops. Assuming an average CO₂ reduction from gasoline use of 25%, ^[39] this would reduce transport CO₂ emissions by 36 Mt.

Furthermore, according to the Beyond the Alternative policy scenario (BAPS), which assumed more energy savings and emission reductions through a set of technological breakthroughs, biofuels use in road transport would double compared to the Alternative policy scenario.

A second IEA report, Energy Technology Perspectives 2006 ( IEA, 2006a [NPR] ), evaluates a series of more ambitious scenarios that yield biomass displacement of 13–25% of transport energy demand by 2050, compared to Baseline levels of 3% displacement. Two scenarios, called Accelerated Technology (ACT) Map and TECH Plus, assume economic incentives equivalent to 25 US$/tCO₂, increased support for research and development, demonstration, and deployment programmes, and policy instruments to overcome commercialization barriers. Both scenarios have optimistic assumptions about the success of efforts to reduce fuel production costs, increase crop yields, and so forth. In the ACT Map scenario, transport biofuels production reaches 480 Mtoe in 2050, accounting for 13% of total transport demand; in TECH Plus, biofuels represents 25% of transport energy demand by 2050 . These displacements yield CO₂ reductions (below the Baseline levels) of 1800 MtCO₂ in Map and 2300 MtCO₂ in TECH Plus, with the major contributors being biodiesel from Fischer Tropsch conversion and ethanol from both sugar crops and cellulosic feedstocks; biodiesel from vegetable oil and ethanol from grains represent somewhat lower shares.

Although the report does not provide quantitative estimates of CO₂ reduction in 2030, it presents qualitative information ( Table 3.5 of the IEA report) that implies that 2030 -levels of biodiesel from vegetable oil and ethanol from grain and sugar crops are similar to 2050 -levels, but biodiesel from Fischer Tropsch conversion, a major source in 2050, plays little role in 2030 and cellulosic ethanol is also significantly lower in 2030 than in 2050 . The implied 2030 potential from the two scenarios appears to be about 600– 1500 MtCO₂.

5.4.2.4 Totals

The estimates discussed above can be summarized as follows:

Light-duty vehicles:

718–766 MtCO₂ at carbon prices less than 100 US$/tCO₂

689–718 MtCO₂ at carbon prices less than 50 US$/tCO₂

669–718 MtCO₂ at carbon prices less than 20 US$/tCO₂

369–697 MtCO₂ at carbon prices less than 0 US$/tCO₂

Aircraft:

150 MtCO2 at carbon prices less than 50 US$/tCO₂

280 MtCO₂ at carbon prices less than 100 US$/tCO₂

Biofuels:

600– 1500 MtCO₂ at carbon prices less than 25 US$/tCO₂

Although presumably the potential for biofuels penetration would be higher above the cited 25 US$/tCO₂ carbon price, the total potential for a carbon price of 100 US$/tCO₂ for the three mitigation sources is about 1600 – 2550 MtCO₂. Much of this potential appears to be located in OECD North America and Europe. Note, however, that the potential is measured as the ‘further’ reduction in CO₂ emissions from the Reference scenario, which assumes that substantial amounts of biofuels will be produced in Brazil and elsewhere and significant improvements in fuel efficiency will occur in China and in other industrializing nations without further policy measures.

5.5.1.1 Land use and transport planning

Energy use for urban transport is determined by a number of factors, including the location of employment and residential locations. In recent decades, most cities have been increasing their dependence on the automobile and decreasing dependence on public transport. In some cases increasing motorization is the result of deliberate planning – what became known as ‘predict and provide’ (The Royal Commission on Transport and the Environment, 1994; ( Goodwin, 1999 ) ). This planning and programming process played a central role in developed countries during the second half of the 20^th century. In many developing countries, the process of motorization and road building is less organized, but is generally following the same motorization path, often at an accelerated rate.

Income plays a central role in explaining motorization. But cities of similar wealth often have very different rates of motorizsation. Mode shares vary dramatically across cities, even within single countries. The share of trips by walking, cycling and public transport is 50% or higher in most Asian, African and Latin American cities, and even in Japan and Western Europe ( Figure 5.17 ). Coordination of land use and transport planning is key to maintaining these high mode shares.

( Kenworthy and Laube 1999 ) ) pointed out that high urban densities are associated with lower levels of car ownership and car use and higher levels of transit use. These densities are decreasing almost everywhere. Perhaps the most important strategy and highest priority to slow motorization is to strengthen local institutions, particularly in urban areas ( Sperling and Salon, 2002 [NPR, SRC] ).

Figure 5.17: Modal split for the cities represented in the Millennium Cities Database for Sustainable Transport by region

Source: Kenworthy & Laube, 2002 [NotFound] .

Some Asian cities with strong governments, especially Hong Kong, Singapore and Shanghai are actively and effectively pursuing strategies to slow motorization by providing high quality public transport and coordinating land use and transport planning ( (Cullinane, 2002; ) ( Willoughby, 2001; ) ( Cameron et al., 2004; ) Sperling and Salon, 2002 [NPR, SRC] ).

There are many other examples of successfully integrated land use and transport planning, including Stockholm and Portland, Oregon (USA) (Abbott, 2002; Lundqvist, 2003 [NPR] ). They mostly couple mixed-use and compact land use development with better public transport access to minimize auto dependence. The effectiveness of these initiatives in reducing sprawl is the subject of debate, especially in the USA ( (Song and Knaap, 2004; ) ( Gordon and Richardson, 1997; ) ( Ewing, 1997 ) ). There are several arguments that the settlement pattern is largely determined, so changes in land use are marginal, or that travel behaviour may be more susceptible to policy interventions than land-use preferences ( Richardson and Bae, 2004 [NPR] ( Ewing and Cervero 2001 ) ) found that typical elasticity of vehicle-km travelled with respect to local density is –0.05, while Pickrell 1999 [NPR] ) noted that reduction in auto use become significant only at densities of 4000 people or more per square kilometre – densities rarely observed in US suburbs, but often reached elsewhere ( Newman and Kenworthy, 1999 [NPR] ). Coordinated transport and land-use methods might have greater benefits in the developing world where dense mixed land use prevails and car ownership rate is low. Curitiba is a prime example of coordinated citywide transport and land-use planning ( (Gilat and Sussman, 2003; ) Cervero, 1998 [NPR] ).

The effectiveness of policies in shifting passengers from cars to buses and rails is uncertain. The literature on elasticity with respect to other prices (cross price elasticity) is not abundant and likely to vary according to the context ( Hensher, 2001 [NPR] ). The Transport Research Laboratory guide showed several cross price elasticity estimates with considerable variance in preceding studies ( (TRL, 2004 ) ( Goodwin 1992 ) ) gave an average cross elasticity of public transport demand with respect to petrol prices of +0.34. ( Jong and Gunn 2001 ) ) also gave an average cross elasticity of public transport trips with respect to fuel price and car time of +0.33 and +0.27 in the short term and +0.07 and +0.15 in the long term.

The literature on mode shifts from cars to new rail services is also limited. A monitoring study of Manchester indicated that about 11% of the passengers on the new light rail would have otherwise used their cars for their trips ( (Mackett and Edwards, 1998 ) ), while a Japanese study of four domestic rails and monorails showed that 10–30% of passengers on these modes were diverted from car mode. The majority of the passengers were transferred from alternative bus and rail routes (Japanese Ministry of Land, Infrastructure and Transport and Institute of Highway Economics, 2004 ). The Transport Research Laboratory guide ( 2004 ) contained international evidence of diversion rates from car to new urban rail ranging from 5–30%. These diversion rates are partly related to car mode share, in the sense that car share is so high in the USA and Australia that ridership on new rail systems is more likely to come from cars in those countries (Booz Allen & Hamilton 1999 [NotFound] , cited in Transport Research Laboratory, 2004 ). It is also known that patronage of metros for cities in the developing world has been drawn almost exclusively from existing public transport users or through generation effects ( (Fouracre et al., 2003 ) ).

The literature suggests that in general, single policies or initiatives tend to have a rather modest effect on the motorization process. The key to restraining motorization is to cluster a number of initiatives and policies, including improved transit service, improved facilities for NMT (Non-motorized transport) and market and regulatory instruments to restrain car ownership and use ( Sperling and Salon, 2002 [NPR, SRC] ). Various pricing and regulatory instruments are addressed below.

Investment appraisal is an important issue in transport planning and policy. The most widely applied appraisal technique in transport is cost benefit analysis (CBA) ( Nijkamp et al., 2003 [NPR] ). In CBA, the cost of CO₂ emissions can be indirectly included in the vehicle operating cost or directly counted at an estimated price, but some form of robustness testing is useful in the latter case. Alternatively, the amount of CO₂ emissions is listed on an appraisal summary table of Multi-Criteria Analysis (MCA) as a part of non-monetized benefits and costs ( Mackie and Nellthorp, 2001 [NPR] ;( Grant-Muller et al., 2001; ) Forkenbrock and Weisbrod, 2001 [NPR] ; Japanese Study Group on Road Investment Evaluation, 2000 ). To the extent that the cost of CO₂ emissions has a relatively important weight in these assessments, investments in unnecessarily carbon-intensive projects might be avoided. Strategic CBA can further make transport planning and policy carbon-efficient by extending CBA to cover multi-modal investment alternatives, while Strategic Environmental Assessment (SEA) can accomplish it by including multi-sector elements. ( ECMT, 2000 [NPR] ; ECMT, 2004b [NPR] ).

5.5.1.2 Taxation and pricing

Transport pricing refers to the collection of measures used to alter market prices by influencing the purchase or use of a vehicle. Typically measures applied to road transport are fuel pricing and taxation, vehicle license/registration fees, annual circulation taxes, tolls and road charges and parking charges. Table 5.14 presents an overview of examples of taxes and pricing measures that have been applied in some developing and developed countries.

Table 5.14: Taxes and pricing in the transport sector in developing and developed countries

Pricing, taxes and charges, apart from raising revenue for governments, are expected to influence travel demand and hence fuel demand and it is on this basis that GHG reduction can be realized.

Transport pricing can offer important gains in social welfare. For the UK, France and Germany together, ( OECD, 2003 [NPR] ) estimates net welfare gains to society of optimal charges (set at the marginal social cost level) at over 20 billion €/yr (22.6 US$/yr).

Although the focus here is on transport pricing options to limit CO₂ emissions, it should be recognized that many projects and policies with that effect are not focused on GHG emissions but rather on other objectives. A pricing policy may well aim simultaneously at reducing local pollution and GHG emissions, accidents, noise and congestion, as well as generating State revenue for enlarging of social welfare and/or infrastructure construction and maintenance. Every benefit with respect to these objectives may then be assessed simultaneously through CBA or MCA; they may be called co-benefits. Governments can take these co-benefits into account when considering the introduction of transport pricing such as for fuel. This is all the more important since a project could be not worth realising if only one particular benefit is considered, whereas it could very well be proved beneficial when adding all the co-benefits.

Taxes

Empirically, throughout the last 30 years, regions with relatively low fuel prices have low fuel economy (USA, Canada, Australia) and regions where relatively high fuel prices apply (due to fuel taxes) have better car fuel economy (Japan and European countries). For example, fuel taxes are about 8 times higher in the UK than in the USA, resulting in fuel prices that are about three times higher. UK vehicles are about twice as fuel-efficient; mileage travelled is about 20% lower and vehicle ownership is lower as well. This also results in lower average per capita fuel expenditures. Clearly, automobile use is sensitive to cost differences in the long run ( VTPI, 2005 [NPR] ). In theory, long run impact of increases in fuel prices on fuel consumption are likely to be about 2 to 3 times greater than short run impact ( VTPI, 2005 [NPR] ). Based on the price elasticities ( (Goodwin et al., 2004 ) ) judged to be the best defined results for developed countries, if the real price of fuel rises by 10% and stays at that level, the volume of fuel consumed by road vehicles will fall by about 2.5% within a year, building up to a reduction of over 6% in the longer run (about 5 years or so), as shown in Table 5.15 .

Table 5.15: Impact of a permanent increase in real fuel prices by 10%

An important reason why a fuel or CO₂ tax would have limited effects is that price elasticities tend to be substantially smaller than the income elasticities of demand. In the long run the income elasticity of demand is a factor 1.5–3 higher than the price elasticity of total transport demand ( (Goodwin et al., 2004 ) ). In developing countries, where incomes are lower, the demand response to price changes may be significantly more elastic.

Recent evidence suggests that the effect of CO₂ taxes and high fuel prices may be having a shrinking effect in the more car-dependent societies. While the evidence is solid that price elasticities indicated in Table 5.15 and used by Goodwin were indeed around –0.25 (i.e., 2.5% reduction in fuel for every 10% increase in price), in earlier years, new evidence indicates a quite different story. Small and van Dender 2007 [2007] ) found that price elasticities in the USA dropped to about –0.11 in the late 1990 s, and Hughes et al. 2006 [NPR, SRC] ) found that they dropped even further in 2001 – 2006, to about –0.04. The explanation seems to be that people in the USA have become so dependent on their vehicles that they have little choice but to adapt to higher prices. One might argue that these are short term elasticities, but the erratic nature of gasoline prices in the USA (and the world) result in drivers never exhibiting long-term behavior. Prices drop before they seriously consider changing work or home locations or even buying more efficient vehicles. If oil prices continue to cycle up and down, as many expect, drivers may continue to cling to their current behaviors. If so, CO₂ taxes would have small and shrinking effects in the USA and other countries where cars are most common.

As an alternative to fuel taxes, registration and circulation taxes can be used to incentivise the purchase (directly) and manufacturing (indirectly) of fuel-efficient cars. This could be done through a revenue neutral fee system, where fuel-efficient cars receive a rebate and guzzler cars are faced with an extra fee. There is evidence that incentives given through registration taxes are more effective than incentives given through annual circulation taxes ( Annema et al., 2001 [NPR] ). Buyers of new cars do not expect to be able to pass on increased registration taxes when selling the vehicle. Due to refunds on registration taxes for cars that were relatively fuel efficient compared to similar sized cars, the percentage of cars sold in the two most fuel efficient classes increased from 0.3%–3.2% (cars over 20% more fuel efficient than average) and from 9.5%–16.1% (for cars between 10 and 20% more fuel efficient than average) in the Netherlands ( ADAC, 2005 [NPR] ). After the abolishment of the refunds, shares decreased again. COWI 2002 [NPR] ) modelled the impact on fuel efficiency of reforming current registration and circulation taxes so they would depend fully on the CO₂ emissions of new cars. Calculated reduction percentages varied from 3.3–8.5% for 9 European countries, depending on their current tax bases.

( Niederberger 2005 ) ) outlines a voluntary agreement with the Swiss government under which the oil industry took responsibility for GHG emissions from the road transport sector, which they supply with fuel. As of 1 October 2005, Swiss oil importers voluntarily contribute the equivalent of about 5 cents per gallon (approx. 80 million US$ annually) into a climate protection fund that is invested via a non-profit (non-governmental) foundation into climate mitigation projects domestically and abroad (via the emerging carbon market mechanisms of the Kyoto Protocol). Cost savings (compared with an incentive tax) are huge and the private sector is in charge of investing the funds effectively. A similar system in the USA could generate 9 billion US$ in funds annually to incentivize clean alternative fuels and energy efficient vehicles, which could lower US dependency on foreign fuel sources. This policy is also credible from a sustainable development perspective than the alternative CO₂ tax, since the high CO₂ tax would have led to large-scale shifts in tank tourism – and bookkeeping GHG reductions for Switzerland – although the real reductions would have been less than half of the total effect and neighbouring countries would have been left with the excess emissions.

Licensing and parking charges

The most renowned area licensing and parking charges scheme has been applied in Singapore with effective reduction in total vehicular traffic and hence energy (petroleum) demand ( Fwa, 2002 [NPR] ). The area licensing scheme in Singapore resulted in 1.043 GJ per day energy savings with private vehicular traffic reducing by 75% ( Fwa, 2002 [NPR] ).

Unfortunately there is currently a lack of data on potential GHG savings associated with policy, institutional and fiscal reforms/measures with respect to transport particularly in other developing countries. General estimates of reduction in use of private vehicle operators resulting from fuel pricing and taxing are 15–20% (World Bank, 2002 [NPR] ; Martin et al., 1995 [NPR] ).

Table 5.16: Potential energy and GHG savings from pricing, taxes and charges for road transport

5.5.1.3 Regulatory and operational measures

Although pricing and fiscal instruments are obvious tools for government policy, they are often not very effective, as reflected by the potential reduction in fuel savings ( IEA, 2003 [NPR] ). Potential effective (and cost-effective) non-fiscal measures that can be effective in an oil crisis are regulatory measures such as:

• Lower speed limits on motorways;
• High occupancy vehicle requirements for certain roads and networks;
• Vehicle maintenance requirements;
• Odd/even number plate and other driving restrictions;
• Providing information on CO₂ emission performances of vehicles (labelling);
• Establishing carbon standards for fuels;
• Direct traffic restrictions (e.g., no entry into business district);
• Free/expanded urban public transport;
• Encouraging alternatives to travel (e.g., greater telecommuting);
• Emergency switching from road to rail freight;
• Reducing congestion through removal of night-time/weekend driving bans for freight.

IEA 2003 [NPR] ) indicates that such measures could contribute to significant oil savings. This is a typical case where a portfolio of measures is applied together and they would work well with adequate systems of monitoring and enforcement.

For the measures to be implemented effectively considerable preparatory work is necessary and Table 5.17 shows examples of what could be done to ensure the measures proposed above can be effective in oil savings.

Table 5.17: Preparations required to implement some regulatory measures

The combined effect of these regulatory measures used to target light-duty vehicles (in addition to blending non-petroleum fuels with gasoline and diesel) is estimated to be a reduction of 15% of daily fuel consumption.

In OECD countries vehicles consume 10–20% more fuel per km than indicated by their rated efficiency. It is estimated that 5–10% reduction in fuel consumption can be achieved by stronger inspection and vehicle maintenance programmes, adoption of on board technologies, more widespread driver training and better enforcement and control of vehicle speeds.

Vehicle travel demand can be reduced by 10–15% by aggressively combining infrastructure improvements, intelligent transport technologies and systems (e.g., better routing systems and congestion reduction), information systems and better transit systems in addition to road pricing.

Another regulatory approach, under consideration in California as part of its 2006 Global Warming Solutions Act, is carbon-based fuel standards. Fuel suppliers would be required to reduce the carbon content of their fuels according to a tightening schedule. For instance, gasoline from conventional oil would be rated at 1.0, ethanol from corn and natural gas at 0.8, electricity for vehicles at 0.6 and so on. The fuel suppliers would be allowed to trade and bank credits and car makers would be required to produce vehicles at an amount that corresponds to the planned sales of alternative fuels. Reductions of 5% or more in transport fuel GHGs by 2020 are envisioned, with much greater reductions in later years.

5.5.1.4 Fuel economy standards – road transport

Most industrialized nations now impose fuel economy requirements (or their equivalent in CO₂ emissions requirements) on new light-duty vehicles ( Plotkin, 2004 [SRC] ; An and Sauer, 2004 [NPR] ). The first standards were imposed by the United States in 1975, requiring 27.5 mpg (8.55 L/100 km) corporate fleet averages for new passenger cars and 20.7 mpg (11.36 L/100 km) for light trucks (based on tests instituted by the US Environmental Protection Agency, using the ‘CAFE’ driving cycle) by 1985 . The passenger car standard remains unchanged, whereas the light truck standard has recently been increased to 22.2 mpg (10.6 L/100 km) for the 2007 model year and to 23.5 mpg (10.0 L/100 km) in model year 2010 . ^[40] Additional standards (some voluntary) include:

• European Union: a 2008 fleet wide requirement ^[41] of 140 gCO₂/km, about 41 mpg (5.74 L/100 km) of gasoline equivalent, using the New European Driving Cycle (NEDC), based on a Voluntary Agreement between the EU and the European manufacturers, with the Korean and Japanese manufacturers following in 2009. Recent slowing of the rate of efficiency improvement has raised doubts that the manufacturers will achieve the 2008 and 2009 targets (Kageson, 2005).
• Japan: a 2010 target of about 35.5 mpg (6.6 L/100 km) for new gasoline passenger vehicles, using the Japan 10/15 driving cycle based on weight-class standards.
• China: weight-class standards that are applied to each new vehicle using the NEDC driving cycle, with target years of 2005 and 2008. At the historical mix of vehicles, the standards are equivalent to fleet targets of about 30.4 mpg (7.7 L/100 km) by 2005 and 32.5 mpg (7.2 L/100 km) by 2008 (An and Sauer, 2004).
• Australia: a 2010 target for new vehicles of 18% reduction in average fuel consumption relative to the 2002 passenger car fleet, corresponding to 6.8 L/100 km, or 34.6 mpg. (DfT, 2003), based on a voluntary agreement between industry and government.
• The State of California has established GHG emission standards for new light-duty vehicles designed to reduce per-vehicle emissions by 22% in 2012 and 30% by 2016. Several US states have decided to adopt these standards, as well. At the time of writing, US industry and the federal government were fighting these standards in the courts.

The NEDC and Japan 10/15 driving cycles are slower than the US CAFE cycle and, for most vehicles (though probably not for hybrids), will yield lower measured fuel economy levels than the CAFE cycle for the same vehicles. Consequently, if they reach their targets, the EU, Japanese and Chinese fleets are likely to achieve fuel economies higher than implied by the values above if measured on the US test. A suggested correction factor (for the undiscounted test results) is 1.13 for the EU and China and 1.35 for Japan ( An and Sauer, 2004 [NPR] ), though these are likely to be at the high end of the possible range of values for such factors. ^[42] Figure 5.18 shows the ‘corrected’ comparison of standards.

Figure 5.18: Fuel economy and GHG emission standards

Note: all the fuel economy targets represent test values based on artificial driving cycles. The standards in the EU and Australia are based on voluntary agreements. In most cases, actual on-road fuel economy values will be lower; for example, the US publishes fuel economy estimates for individual LDVs that are about 15% lower than the test val-ues and even these values appear to be optimistic. Miles/gallon is per US gallon.

Recent studies of the costs and fuel savings potential of technology improvements indicate considerable opportunity to achieve further fleet fuel economy gains from more stringent standards. For example, the US National Research Council ( NRC, 2002 [NPR] ) estimates that US light-duty vehicle fuel economy can be increased by 25–33% within 15 years with existing technologies that cost less than the value of fuel saved. A study by Ricardo Consulting Engineers for the UK Department for Transport ( Owen and Gordon, 2002 [NPR] ) develops a step-wise series of improvements in a baseline diesel passenger car that yields a 38% reduction in CO₂ emissions (a 61% increase in fuel economy), to 92 g/km, by 2013 using parallel hybrid technology at an incremental cost of 2300 –3,100 £ (4200–5700 US$) with a 15,300 £ (28,000 US$) baseline vehicle. Even where fuel savings will outweigh the cost of new technologies, however, the market will not necessarily adopt these technologies by itself (or achieve the maximum fuel economy benefits from the technologies even if they are adopted). Two crucial deterrents are, first, that the buyers of new vehicles tend to consider only the first three years or so of fuel savings ( NRC, 2002 [NPR] ; Annema et al., 2001 [NPR] ), and second, that vehicle buyers will take some of the benefits of the technologies in higher power and greater size rather than in improved fuel economy. Further, potential benefits for consumers over the vehicle’s lifetime are generally small, while risks for producers are high ( Greene, 2005 [NPR, MoS, SRC] ). Also, neither the purchasers of new vehicles nor their manufacturers will take into account the climate effects of the vehicles.

Strong criticisms have been raised about fuel economy standards, particularly concerning claimed adverse safety implications of weight reductions supposedly demanded by higher standards and increased driving caused by the lower fuel costs (per mile or km) associated with higher fuel economy.

The safety debate is complex and not easily summarized. Although there is no doubt that adding weight to a vehicle improves its safety in some types of crashes, it does so at the expense of other vehicles; further, heavy light trucks have been shown to be no safer, and in some cases less safe than lighter passenger cars, primarily because of their high rollover risk ( Ross et al., 2006 [NPR] ). The US National Highway Traffic Safety Administration (NHTSA) has claimed that fleet wide weight reductions ‘reduce’ fleet safety ( Kahane, 2003 [NPR, MoS] ), but this conclusion is strongly disputed ( DRI, 2004 [NPR] ; NRC, 2002 [NPR] ). An important concern with the NHTSA analysis is that it does not separate the effects of vehicle weight and size. In any case, other factors, e.g., overall vehicle design and safety equipment, driver characteristics, road design, speed limits and alcohol regulation and enforcement play a more significant role in vehicle safety than does average weight.

Some have argued that increases in driving associated with reduced fuel cost per mile will nullify the benefits of fuel economy regulations. Increased driving ‘is’ likely, but it will be modest and decline with higher income and increased motorization. Recent data implies that a driving ‘rebound’ would reduce the GHG reduction (and reduce oil consumption) benefits from higher standards by about 10% in the United States ( Small and van Dender, 2007 [2007] ) but more than this in less wealthy and less motorized countries.

In deciding to institute a new fuel economy standard, governments should consider the following:

• Basing stringency decisions on existing standards elsewhere requires careful consideration of differences between the home market and compared markets in fuel quality and availability; fuel economy testing methods; types and sizes of vehicles sold; road conditions that may affect the robustness of key technologies; and conditions that may affect the availability of technologies, for example, availability of sophisticated repair facilities.
• There are a number of different approaches to selecting stringency levels for new standards. Japan selected its weight class standards by examining ‘top runners’ – exemplary vehicles in each weight class that could serve as viable targets for future fleet wide improvements. Another approach is to examine the costs and fuel saving effects of packages of available technologies on several typical vehicles, applying the results to the new vehicle fleet (NRC, 2002). Other analyses have derived cost curves (percent increase in fuel economy compared with technology cost) for available technology and applied these to corporate or national fleets (Plotkin et al., 2002). These approaches are not technology-forcing, since they focus on technologies that have already entered the fleet in mass-market form. More ambitious standards could demand the introduction of emerging technologies. Selection of the appropriate level of stringency depends, of course, on national goals and concerns. Further, the selection of enforcement deadlines should account for limitations on the speed with which vehicle manufacturers can redesign multiple models and introduce the new models on a schedule that avoids severe economic disruption.
• The structure of the standard is as important as its level of stringency. Basing target fuel economy on vehicle weight (Japan, China) or engine size (Taiwan, South Korea) will tend to even out the degree of difficulty the standards impose on competing automakers, but will reduce the potential fuel economy gains that can be expected (because weight-based standards eliminate weight reduction and engine-size-based standards eliminate engine downsizing as viable means of achieving the standards). Basing the standard on vehicle wheelbase times track width may provide safety benefits by providing a positive incentive to maintain or increase these attributes. Using a uniform standard for all vehicles or for large classes of vehicles (as in the US) is simple and easy to explain, but creates quite different challenges on different manufacturers depending on the market segments they focus on.
• Allowing trading of fuel economy ‘credits’ among different vehicles or vehicle categories in an automaker’s fleet, or even among competing automakers, will reduce the overall cost of standards without reducing the total societal benefits, but may incur political costs from accusations of allowing companies or individuals to ‘buy their way out’ of efficiency requirements.
• Alternatives (or additions) to standards are worth investigating. For example, ‘feebates’, which award cash rebates to new vehicles whose fuel economy is above a designated level (often the fleet average) and charge a fee to vehicles with lower fuel economy, may be an effective market-based measure to increase fleet fuel economy. An important advantage of feebates is that they provide a ‘continuous’ incentive to improve fuel economy, because an automaker can always gain a market advantage by introducing vehicles that are more efficient than the current average.

5.5.1.5 Transport Demand Management

Transport Demand Management (TDM) is a formal designation for programmes in many countries that improve performance of roads by reducing traffic volumes ( (Litman, 2003 ) ). There are many potential TDM strategies in these programmes with a variety of impacts. Some improve transport diversity (the travel options available to users). Others provide incentives for users to reduce driving, changing the frequency, mode, destination, route or timing of their travel. Some reduce the need for physical travel through mobility substitutes or more efficient land use. Some involve policy reforms to correct current distortions in transport planning practices. TDM is particularly appropriate in developing country cities, because of its low costs, multiple benefits and potential to redirect the motorization process. In many cases, effective TDM during early stages of development can avoid problems that would result if communities become too automobile dependent. This can help support a developing country’s economic, social and environmental objectives ( Gwilliam et al., 2004 [NPR] ).

The set of strategies to be implemented will vary depending on each country’s demographic, geographic and political conditions. TDM strategies can have cumulative and synergetic impacts, so it is important to evaluate a set of TDM programmes as a package, rather than as an individual programme. Effective strategies usually include a combination of positive incentives to use alternative modes (‘carrots’ or ‘sweeteners’) and negative incentives to discourage driving (‘sticks’ or ‘levellers’). Recent literature gives a comprehensive overview of these programmes with several case studies (May et al., 2003; ( Litman, 2003; ) WCTRS and IPTS, 2004 [NotFound] ). Some major strategies such as pricing and land-use planning are addressed above. Below is a selective review of additional TDM strategies with significant potential to reduce vehicle travel and GHGs.

Employer travel reduction strategies gained prominence from a late 1980 s regulation in southern California that required employers with 100 or more employees to adopt incentives and rules to reduce the number of car trips by employees commuting to work ( (Giuliano et al., 1993 ) ). The State of Washington in the USA kept a state law requiring travel plans in its most urban areas for employers with 100 or more staff. The law reduced the percentage of employees in the targeted organizations who drove to work from 72–68% and affected about 12% of all trips made in the area. In the Netherlands, the reduction in single occupant commute trips from a travel plan averaged 5–15%. In the UK, in very broad terms, the average effectiveness of UK travel plans might be 6% in trips by drive alone to work and 0.74% in the total vehicle-km travelled to work by car. The overall effectiveness was critically dependent on both individual effectiveness and levels of plan take-up ( (Rye, 2002 ) ).

Parking supply for employees is so expensive that employers naturally have an incentive to reduce parking demand. The literature found the price elasticity of parking demand for commuting at –0.31 to –0.58 ( Deuker et al., 1998 [NPR] ) and –0.3 ( Veca and Kuzmyak, 2005 [NPR] ) based on a non-zero initial parking price. The State of California enacted legislation that required employers with 50 or more persons who provided parking subsidies to offer employees the option to choose cash in lieu of a leased parking space, in a so-called parking cash-out programme. In eight case studies of employers who complied with the cash-out programme, the solo driver share fell from 76% before cashing out to 63% after cashing out, leading to the reduction in vehicle-km for commuting by 12%. If all the commuters who park free in easily cashed-out parking spaces were offered the cash option in the USA, it would reduce vehicle-km travelled per year by 6.3 billion ( (Shoup, 1997 ) ).

Reducing car travel or CO₂ emissions by substituting telecommuting for actual commuting has often been cited in the literature, but the empirical results are limited. In the USA, a micro-scale study estimated that 1.5% of the total workforce telecommuted on any day, eliminating at most 1% of total household vehicle-km travelled ( Mokhtarian, 1998 [MoS] ), while a macro-scale study suggested that telecommuting reduced annual vehicle-km by 0–2% ( (Choo et al., 2005 ) ).

Reduction of CO₂ emissions by hard measures, such as car restraint, often faces public opposition even when the proposed measures prove effective. Soft measures, such as a provision of information and use of communication strategies and educational techniques ( OECD, 2004a [NPR] ) can be used for supporting the promotion of hard measures. Soft measures can also be directly helpful in encouraging a change in personal behaviour leading to an efficient driving style and reduction in the use of the car ( Jones, 2004 [NPR] ). Well organized soft measures were found to be effective for reducing car travel while maintaining a low cost. Following travel awareness campaigns in the UK, the concept of Individualized marketing, a programme based on a targeted, personalized, customized marketing approach, was developed and applied in several cities for reducing the use of the car. The programme reduced car trips by 14% in an Australian city, 12% in a German city and 13% in a Swedish city. The Travel Blending technique was a similar programme based on four special kits for giving travel-feedback to the participants. This programme reduced vehicle-km travelled by 11% in an Australian city. The monitoring study after the programme implementation in Australian cities also showed that the reduction in car travel was maintained ( Brog et al., 2004 [NPR] ;( Taylor and Ampt, 2003 ) ). Japanese cases of travel-feedback programmes supported the effectiveness of soft measures for reducing car travel. The summary of the travel-feedback programmes in residential areas, workplaces and schools indicated that car use was reduced by 12% and CO₂ emissions by 19%. It also implied that the travel-feedback programmes with a behavioural plan requiring a participant to make a plan for a change showed better results than programmes without one ( Fujii and Taniguchi, 2005 [NPR] ).

5.5.2.1 Aviation

IPCC 1999 [NPR] ), ICAO/FESG ( 2004a, b), Wit et al. 2002 [NPR, SRC] and 2005 Cames and Deuber 2004 [NPR] Arthur Andersen 2001 [NotFound] ) and others have examined potential economic instruments for mitigating climate effects from aviation.

At the global level no support exists for the introduction of kerosene taxes. The ICAO policy on exemption of aviation fuel from taxation has been called into question mainly in European states that impose taxes on fuel used by other transport modes and other sources of GHGs. A study by Resource Analysis 1999 [NPR] ) shows that introducing a charge or tax on aviation fuel at a ‘regional’ level for international flights would give rise to considerable distortions in competition and may need amendment of bilateral air service agreements. In addition, the effectiveness of a kerosene tax imposed on a regional scale would be reduced as airlines could take ‘untaxed’ fuel onboard into the taxed area (the so-called tankering effect).

Wit and Dings 2002 [NPR, SRC] ) analyzed the economic and environmental impacts of en-route emission charges for all flights in European airspace. Using a scenario-based approach and an assumed charge level of 50 US$/tCO₂, the study found a cut in forecast aviation CO₂ emissions in EU airspace of about 11 Mt (9%) in 2010 . This result would accrue partly (50%) from technical and operational measures by airlines and partly from reduced air transport demand. The study found also that an en-route emission charge in European airspace designed in a non-discriminative manner would have no significant impact on competition between European and non-European carriers.

In a study prepared for CAEP/6, the Forecasting and Economic Analysis Support Group (ICAO/FESG, 2004a ) considered the potential economic and environmental impacts of various charges and emission trading schemes. For the period 1998 – 2010, the effects of a global CO₂ charge with a levy equivalent to 0.02 US$/kg to 0.50 US$/kg jet fuel show a reduction in global CO₂ emissions of 1–18%. This effect is mainly caused by demand effects (75%). The AERO modelling system was used to conduct the analyses ( Pulles, 2002 [NPR] ).

As part of the analysis of open emission trading systems for CAEP/6, an impact assessment was made of different emission trading systems identified in ICF et al. 2004 [NPR] ). The ICAO/FESG report ( 2004b ) showed that under a Cap-and-Trade system for aviation, total air transport demand will be reduced by about 1% compared to a base case scenario FESG,) showed that under a Cap-and-Trade system for aviation, total air transport demand will be reduced by about 1% compared to a base case scenario FESG,). In this calculation, a 2010 target of 95% of the 1990 -level was assumed for aviation on routes from and to Annex-I countries and the more developed non-Annex-I countries such as China, Hong Kong, Thailand, Singapore, Korea and Brazil. Furthermore a permit price of 20 US$/tCO₂ was assumed. Given the relative high abatement costs in the aviation sector, this scenario would imply that the aviation sector would buy permits from other sectors for about 3.3 billion US$.

In view of the difficulty of reaching global consensus on mitigation policies to reduce GHG emissions from international aviation, the European Commission decided to prepare climate policies for aviation. On 20 December 2006 the European Commission presented a legislative proposal that brings aviation emissions into the existing EU Emissions Trading Scheme (EU ETS). The proposed directive will cover emissions from flights within the EU from 2011 and all flights to and from EU airports from 2012 . Both EU and foreign aircraft operators would be covered. The environmental impact of the proposal may be significant because aviation emissions, which are currently growing rapidly, will be capped at their average level in 2004 – 2006 . By 2020 it is estimated by model analysis that a total of 183 MtCO₂ will be reduced per year on the flights covered, a 46% reduction compared with business-as-usual. However, aviation reduces the bulk of this amount through purchasing allowances from other sectors and through additional supply of Joint Implementation and Clean Development Mechanism credits. In 2020 aviation reduces its own emissions by 3% below business-as-usual ( EC, 2006 [NPR] ).

If emission trading or emission charges were applied to the aviation sector in isolation, the two instruments would in principle be equivalent in terms of cost-effectiveness. However, combining the reduction target for aviation with the emission trading scheme of other sectors increases overall economic efficiency by allowing the same amount of reductions to be made at a lower overall cost to society. Therefore, if aviation were to achieve the same environmental goal under emission trading and emission charges, the economic costs for the sector and for the economy as a whole would be lower if this was done under an emission trading scheme including other sectors rather than under a charging system for aviation only.

Alternative policy instruments that may be considered are voluntary measures or fuel taxation for domestic flights. Fuel for domestic flights, which are less vulnerable to economic distortions, is already taxed in countries such as the USA, Japan, India and the Netherlands. In parallel to the introduction of economic instruments such as emission trading, governments could improve air traffic management.

Policies to address the full climate impact of aviation

A major difficulty in developing a mitigation policy for the climate impacts of aviation is how to cover non-CO₂ climate impacts, such as the emission of nitrogen oxides (NO_x) and the formation of condensation trails and cirrus clouds (see also Box 5.1 in Section 5.2 IPCC 1999 [NPR] ) estimated these effects to be about 2 to 4 times greater than those of CO₂ alone, even without considering the potential impact of cirrus cloud enhancement. This means that the perceived environmental effectiveness of any mitigation policy will depend on the extent to which these non-CO₂ climate effects are also taken into account.

Different approaches may be considered to account for non-CO₂ climate impacts from aviation ( Wit et al., 2005 [NPR, SRC] ). A first possible approach is where initially only CO₂ from aviation is included in for example an emission trading system, but flanking instruments are implemented in parallel such as differentiation of airport charges according to NO_x emissions.

Another possible approach is, in case of emission trading for aviation, a requirement to surrender a number of emission permits corresponding to its CO₂ emissions multiplied by a precautionary average factor reflecting the climate impacts of non-CO₂ impacts. It should be emphasised that the metric that is a suitable candidate for incorporating the non-CO₂ climate impacts of aviation in a single metric that can be used as a multiplier requires further development, being fairly theoretical at present. The feasibility of arriving at operational methodologies for addressing the full climate impact of aviation depends not only on improving scientific understanding of non-CO₂ impacts, but also on the potential for measuring or calculating these impacts on individual flights.

5.5.2.2 Shipping

CO₂ emission indexing scheme

The International Maritime Organisation (IMO), a specialized UN agency, has adopted a strategy with regard to policies and measures, focusing mainly on further development of a CO₂ emission indexing scheme for ships and further evaluation of technical, operational and market-based solutions.

The basic idea behind a CO₂ emission index is that it describes the CO₂ efficiency (i.e., the fuel efficiency) of a ship, i.e., the CO₂ emission per tonne cargo per nautical mile. This index could, in the future, assess both the technical features (e.g., hull design) and operational features of the ship (e.g., speed).

In June 2005, at the 53^rd session of the Marine Environment Protection Committee of IMO ( IMO, 2005 [NPR] ), interim guidelines for voluntary ship CO₂ emission indexing for use in trials were approved. The Interim Guidelines should be used to establish a common approach for trials on voluntary CO₂ emission indexing, which enable shipowners to evaluate the performance of their fleet with regard to CO₂ emissions. The indexing scheme will also provide useful information on a ship’s performance with regard to fuel efficiency and may thus be used for benchmarking purposes. The interim guidelines will later be updated, taking into account experience from new trials as reported by industry, organisations and administrations.

A number of hurdles have to be overcome before such a system could become operational. The main bottleneck appears to be that there is major variation in the fuel efficiency of similar ships, which is not yet well understood ( Wit et al., 2004 [NPR, SRC] ). This is illustrated by research by the German delegation of IMO’s Working Group on GHG emission reduction ( IMO, 2004 [NPR] ), in which the specific energy efficiency (i.e., a CO₂ emission index) was calculated for a range of container ships, taking into account engine design factors rather than operational data. The results of this study show that there is considerable scatter in the specific engine efficiency of the ships investigated, which could not be properly explained by the deadweight of the ships, year of build, ship speed and several other ship design characteristics. The paper therefore concludes that the design of any CO₂ indexing scheme and its differentiation according to ship type and characteristics, requires in-depth investigation. Before such a system can be used in an incentive scheme, the reasons for the data scatter need to be understood. This is a prerequisite for reliable prediction of the economic, competitive and environmental effects of any incentive based on this method.

Voluntary use and reporting results of CO₂ emission indexing may not directly result in GHG emission reductions, although it may well raise awareness and trigger certain initial moves towards ‘self regulation’. It might also be a first step in the process of designing and implementing some of the other policy options. Reporting of the results of CO₂ emission indexing could thus generate a significant impetus to the further development and implementation of this index, since it would lead to widespread experience with the CO₂ indexing methodology, including reporting procedure and monitoring, for shipping companies as well as for administrations of states.

In the longer term, in order to be more effective, governments may consider using CO₂ indexing via the following paths:

1. The indexing of ship operational performance is introduced as a voluntary measure and over time developed and adopted as a standard;

2. Based on the experience with the standard, it will act as a new functional requirement when new buildings are ordered, hence over time the operational index will affect the requirements from ship owners related to the energy efficiency of new ships;

3. Differentiation of en route emission charges or existing port dues on the basis of a CO₂ index performance;

4. To use the CO₂ index of specific ship categories as a baseline in a (voluntary) baseline-and-credit programme.

Economic instruments for international shipping

There are currently only a few cases of countries or ports introducing economic instruments to create incentives to reduce shipping emissions. Examples include environmentally differentiated fairway dues in Sweden, the Green Award scheme ^[43] in place in 35 ports around the world, the Green Shipping bonus in Hamburg and environmental differentiation of tonnage tax in Norway. None of these incentives are based on GHG emissions, but generally relate to fuel sulphur content, engine emissions (mainly NO_x), ship safety features and management quality.

Harrison et al. 2004 [NPR] ) explored the feasibility of a broad range of market-based approaches to regulate atmospheric emissions from seagoing ship in EU sea areas. The study focused primarily on policies to reduce the air pollutants SO₂ and NO_x, but the approaches adopted may to a certain extent also be applicable to other emissions, including CO₂. According to a follow-up study by Harrison et al. 2005 [NPR] ) the main obstacles to a programme of voluntary port dues differentiation are to provide an adequate level of incentive, alleviating ports’ competitive concerns and reconciling differentiation with specially negotiated charges. Swedish experience suggests that when combined with a centrally determined mandatory charging programme, these problems may be surmountable. However, in many cases a voluntary system would not likely be viable and other approaches to emissions reductions may therefore be required.

An alternative economic instrument, such as a fuel tax is vulnerable to evasion; that is ships may avoid the tax by taking fuel on board outside the taxed area. Offshore bunker supply is already common practice to avoid paying port fees or being constrained by loading limits in ports. Thus even a global fuel tax could be hard to implement to avoid evasion, as an authority at the port state level would have to collect the tax ( ECON, 2003 [NPR] ). A CO₂-based route charge or a (global) sectoral emission trading scheme would overcome this problem if monitoring is based on the carbon content of actual fuel consumption on a single journey. As yet there is no international literature that analyzes the latter two policy options. Governments may therefore consider investigating the feasibility and effectiveness of emission charges and emission trading as policy instruments to reduce GHG emissions from international shipping.