The adoption of the 2030 Agenda for Sustainable Development by the United Nations (UN) has renewed international efforts to pursue and accurately measure global actions towards sustainable development (United Nations 2015). The 17 SDGs set out the overall goals that are further specified by 169 targets and 232 SDG indicators, many of which relate to transport (United Nations 2017; Lisowski et al. 2020). A sustainable transport system provides safe, inclusive, affordable, and clean passenger and freight mobility for current and future generations (Williams 2017; Litman 2021) so transport is particularly linked to SDGs 3, 7, 8, 9, 11, 12, and 13 (Move Humanity 2018; IRP 2019; WBA 2019; SLoCaT 2019; Yin 2019). Table 10.1 summarises transport-related topics for these SDGs and corresponding research. Section 17.3.3.7 also provides a cross-sectoral overview of synergies and trade-offs between climate change mitigation and the SDGs.

Table 10.1 | Transport and the Sustainable Development Goals: Synergies and trade-offs.

The transport sector directly emitted around 8.9 Gtonnes (Gt) of carbon dioxide equivalent (CO2-eq) in 2019, up from 5.1 GtCO2-eq in 1990 (Figure 10.1). Global transport was the fourth largest source of GHG emissions in 2019 following the power, industry, and the agriculture, forestry and land use (AFOLU) sectors. In absolute terms, the transport sector accounts for roughly 15% of total GHG emissions and about 23% of global energy-related CO2 emissions (IEA 2020a). Transport-related GHG emissions have increased fast over the last two decades, and since 2010, the sector’s emissions have increased faster than for any other end-use sector, averaging +1.8% annual growth (Section 10.7). Addressing emissions from transport is crucial for GHG mitigation strategies across many countries, as the sector represents the largest energy consuming sector in 40% of countries worldwide. In most remaining countries, transport is the second largest energy-consuming sector, reflecting different levels of urbanisation and land use patterns, speed of demographic changes and socio-economic development (IEA 2012; Gota et al. 2019; Hasan et al. 2019; Xie et al. 2019).

As of 2019, the largest source of transport emissions is the movement of passengers and freight in road transport (6.1 GtCO2-eq, 69% of the sector’s total). International shipping is the second largest emission source, contributing 0.8 GtCO2- eq (9% of the sector’s total), and international aviation is third with 0.6 GtCO2-eq (7% of the sector’s total). All other transport emissions sources, including rail, have been relatively trivial in comparison, totalling 1.4 GtCO2- eq in 2019. Between 2010 and 2019, international aviation had among the fastest growing GHG emissions among all segments (+3.4% per year), while road transport remained one of the fastest growing (+1.7% per year) among all global energy-using sectors. Note that the COVID-19-induced economic lockdowns implemented since 2020 have had a very substantial impact on transport emissions – higher than any other sector (Chapter 2). Preliminary estimates from Crippa et al. (2021) suggest that global transport CO2 emissions declined to 7.6 GtCO2 in 2020, a reduction of 11.6% compared to 2019 (Crippa et al. 2021; Minx et al. 2021). These lockdowns affected all transport segments, and particularly international aviation (estimated 45% reduction in 2020 global CO2 emissions), road transport (–10%), and domestic aviation (–9.3%). By comparison, aggregate CO2 emissions across all sectors are estimated to have declined by 5.1% as a result of the COVID-19 pandemic (Section 2.2.2).

Growth in transport-related GHG emissions has taken place across most world regions (see Figure 10.1b). Between 1990 and 2019, growth in emissions was relatively slow in Europe, Australia, Japan and New Zealand, Eurasia, and North America while it was unprecedently fast in other regions. Driven by economic and population growth, the annual growth rates in Eastern Asia, Southern Asia, South-East Asia and Pacific, and Africa were 6.1%, 5.2%, 4.7%, and 4.1%, respectively. Latin America and the Middle East have seen somewhat slower growth in transport-related GHG emissions (annual growth rates of 2.4% and 3.3%, respectively) (ITF 2019; Minx et al. 2021). Section 10.7 provides a more detailed comparison of global transport emissions trends with those from regional and sub-sectoral studies.

The rapid growth in global transport emissions is primarily a result of the fast growth in global transport activity levels, which grew by 73% between 2000 and 2018. Passenger and freight activity growth have outpaced energy efficiency and fuel economy improvements in this period (ITF 2019). The global increase in passenger travel activities has taken place almost entirely in non-OECD countries, often starting from low motorisation rates (SLoCaT 2018a). Passenger cars, two- and three-wheelers, and mini buses contribute about 75% of passenger transport-related CO2 emissions, while collective transport services (bus and railways) generate about 7% of the passenger transport-related CO2 emissions despite covering a fifth of passenger transport globally (Rodrigue 2017; Halim et al. 2018; Sheng et al. 2018; SLoCaT 2018a; Gota et al. 2019). While alternative lighter powertrains have great potential for mitigating GHG emissions from cars, the trend has been towards increasing vehicle size and engine power within all vehicle size classes, driven by consumer preferences towards larger sport utility vehicles (SUVs) (IEA 2020a). On a global scale, SUV sales have been constantly growing in the last decade, with 40% of the vehicles sold in 2019 being SUVs (IEA 2020a) (Section 10.4, Box 10.3).

Indirect emissions from electricity and heat shown in Figure 10.1 account for only a small fraction of current emissions from the transport sector (2%) and are associated with electrification of certain modes like rail or bus transport (Lamb et al. 2021). Increasing transport electrification will affect indirect emissions, especially where carbon-intense electricity grids operate.

Global freight transport, measured in tonne-kilometres (tkm), grew by 68% between 2000 and 2015 and is projected to grow 3.3 times by 2050 (ITF 2019). If unchecked, this growth will make decarbonisation of freight transport very difficult (McKinnon 2018; ITF 2019). International trade and global supply chains from industries frequently involving large geographical distances are responsible for the fast increase of CO2 emissions from freight transport (Yeh et al. 2017; McKinnon 2018), which are growing faster than emissions from passenger transport (Lamb et al. 2021). Heavy-duty vehicles (HDVs) make a disproportionate contribution to air pollution, relative to their global numbers, because of their substantial emissions of particulate matter and of black carbon with high short-term warming potentials (Anenberg et al. 2019).

On-road passenger and freight vehicles dominate global transport-related CO2 emissions and offer the largest mitigation potential (Taptich et al. 2016; Halim et al. 2018). This chapter examines a wide range of possible transport emission reduction strategies. These strategies can be categorised under the ‘Avoid-Shift-Improve’ (ASI) framework described in Chapter 5 (Taptich et al. 2016). ‘Avoid’ strategies reduce total vehicle travel. They include compact communities and other policies that minimise travel distances and promote efficient transport through pricing and demand management programmes. ‘Shift’ strategies shift travel from higher-emitting to lower-emitting modes. These strategies include more multimodal planning that improves active and collective transport modes, complete streets roadway design, high occupant vehicle priority strategies that favour shared modes, Mobility as a Service (MaaS), and multimodal navigation and payment apps. ‘Improve’ strategies reduce per-kilometre emission rates. These strategies include hybrid and electric vehicle incentives, lower-carbon and cleaner fuels, high-emitting vehicle scrappage programmes, and efficient driving and anti-idling campaigns (Lutsey and Sperling 2012; Gota et al. 2015). These topics are assessed within the rest of this chapter, including how combinations of ASI with new technologies can potentially lead from incremental interventions into low-carbon transformative transport improvements that include social and equity benefits (Section 10.8).

Climate change impacts such as extremely high temperatures, intense rainfall leading to flooding, more intense winds and/or storms, and sea level rise can seriously impact transport infrastructure, operations, and mobility for road, rail, shipping, and aviation. Studies since AR5 confirm that serious challenges to all transport infrastructures are increasing, with consequent delays or derailing (Miao et al. 2018; Moretti and Loprencipe 2018; Pérez-Morales et al. 2019; Palin et al. 2021). These impacts have been increasingly documented but, according to Forzieri et al. (2018), little is known about the risks of multiple climate extremes on critical infrastructures at local to continental scales. All roads, bridges, rail systems, and ports are likely to be affected to some extent. Flexible pavements are particularly vulnerable to extreme high temperatures that can cause permanent deformation and crumbling of asphalt (Underwood et al. 2017; Qiao et al. 2019). Rail systems are also vulnerable, with a variety of hazards, both meteorological and non-meteorological, affecting railway asset lifetimes. Severe impacts on railway infrastructure and operations can arise from the occurrence of temperatures below freezing, excess precipitation, storms and wildfires (Thaduri et al. 2020; Palin et al. 2021) as can impacts on underground transport systems (Forero-Ortiz et al. 2020).

Most countries are examining opportunities for combined mitigation-adaptation efforts, using the need to mitigate climate change through transport-related GHG emissions reductions and reduction of pollutants as the basis for adaptation action (Thornbush et al. 2013; Wang et al. 2020). For example, urban sprawl indirectly affects climate processes, increasing emissions and vulnerability, which worsens the potential to adapt (Congedo and Munafò 2014; Macchi and Tiepolo 2014). Hence, using a range of forms of rapid transit as structuring elements for urban growth can mitigate climate change-related risks as well as emissions, reducing impacts on new infrastructure, often in more vulnerable areas (Newman et al. 2017). Such changes are increasingly seen as having economic benefit (Ha et al. 2017), especially in developing nations (Chang 2016; Monioudi et al. 2018).

Since AR5 there has been a growing awareness of the potential and actual impacts from global sea level rise due to climate change on transport systems (Dawson et al. 2016; Rasmussen et al. 2018; IPCC 2019; Noland et al. 2019), particularly on port facilities (Stephenson et al. 2018; Yang et al. 2018b; Pérez-Morales et al. 2019). Similarly, recent studies suggest changes in global jet streams could affect the aviation sector (Staples et al. 2018; Becken and Shuker 2019), and extreme weather conditions can affect runways (heat buckling) and aircraft lift. Combined, climate impacts on aviation could result in payload restrictions and disruptions (Coffel et al. 2017; Monioudi et al. 2018). According to Williams (2017), studies have indicated that the amount of moderate-or-greater clear-air turbulence on transatlantic flight routes in winter will increase significantly in the future as the climate changes. More research is needed to fully understand climate-induced risks to transportation systems.

Available evidence suggests that transport-related CO2 emissions would need to be restricted to about 2 to 3 Gt in 2050 (1.5°C scenario-1.5DS, B2DS), or about 70 to 80% below 2015 levels, to meet the goals set in the Paris Agreement. It also indicates that a balanced and inter-modal application of Avoid, Shift, and Improve measures is capable of yielding an estimated reduction in transport emissions of 2.39 GtCO2-equivalent by 2030 and 5.74 GtCO2-equivalent by 2050 (IPCC 2018; Gota et al. 2019). Such a transformative decarbonisation of the global transport system requires, in addition to technological changes, a paradigm shift that ensures prioritisation of high-accessibility transport solutions that minimise the amount of mobility required to meet people’s needs, and favours transit and active transport modes (Lee and Handy 2018; SLoCaT 2021). These changes are sometimes called disruptive as they are frequently surprising in how they accelerate through a technological system.

The assessment of transport innovations and their mitigation potentials is at the core of how this chapter examines the possibilities for changing transport-related GHG trajectories. The transport technology innovation literature analysed in this chapter emphasises how a mixture of mitigation technology options and social changes are now converging and how, in combination, they may have potential to accelerate trends toward a low-carbon transport transition. Such changes are considered disruptive or transformative (Sprei 2018). Of the current transport trends covered in the literature, this chapter focuses on three key technology and policy areas: electro-mobility in land-based transport vehicles, new fuels for ships and planes, and overall demand reductions and efficiency. These strategies are seen as being necessary to integrate at all levels of governance and, in combination with the creation of fast, extensive, and affordable multimodal public transport networks, can help achieve multiple advantages in accordance with SDGs

Electrification of passenger transport in light-duty vehicles (LDVs) is well underway as a commercial process with socio-technical transformative potential and will be examined in detail in Sections 10.3 and 10.4. But the rapid mainstreaming of electric vehicles (EVs) will still need enabling conditions for land transport to achieve the shift away from petroleum fuels, as outlined in Chapter 3 and detailed in Section 10.8. The other mitigation options reviewed in this chapter are so far only incremental and are less commercial, especially shipping and aviation fuels, so stronger enabling conditions are likely , as detailed further in Sections 10.5 to 10.8. The enabling conditions that would be needed for the development of an emerging technological solution for such fuels are likely to be very different from those for electromobility, but nevertheless they both will need demand and efficiency changes to ensure they are equitable and inclusive.

Section 10.2 sets out the transformation of transport through examining systemic changes that affect demand for transport services and the efficiency of the system. Section 10.3 looks at the most promising technological innovations in vehicles and fuels. The next three sections (10.4, 10.5, and 10.6) examine mitigation options for land transport, aviation, and shipping. Section 10.7 describes the space of solutions assessed in a range of integrated modelling and sectoral transport scenarios. Finally, Section 10.8 sets out what would be needed for the most transformative scenario that can manage to achieve the broad goals set out in Chapter 3 and the transport goals set out in Section 10.7.

The physical characteristics that make up built areas define the urban form. These physical characteristics include the shape, size, density, and configuration of the human settlements. Urban form is intrinsically coupled with the infrastructure that allows human settlements to operate. In the context of the transport sector, urban form and urban infrastructure influence the time and cost of travel, which, in turn, drive travel demand and modal choice (Marchetti and Ausubel 2001; Newman and Kenworthy 2015).

Throughout history, three main urban fabrics have developed, each with different effects on transport patterns based on a fixed travel time budget of around one hour (Newman et al. 2016). The high-density urban fabric developed over the past several millennia favoured walking and active transport for only a few kilometres (km). In the mid-19th century, urban settlements developed a medium-density fabric that favoured trains and trams traveling over 10 to 30 km corridors. Finally, since the mid-20th century, urban form has favoured automobile travel, enabling mass movement between 50 and 60 km. Table 10.2 describes the effect of these urban fabrics on GHG emissions and other well-being indicators.

Table 10.2 | The systemic effect of city form and transport emissions.

Source: Newman et al. (2016); Thomson and Newman (2018); Seto et al. (2021).

Since AR5, urban design has increasingly been seen as a major way to influence the GHG emissions from urban transport systems. Indeed, research suggests that implementing urban form changes could reduce GHG emissions from urban transport by 25% in 2050, compared with a business-as-usual scenario (Creutzig et al. 2015b; Creutzig 2016). Researchers have identified a variety of variables to study the relationship between urban form and transport-related GHG emissions. Three notable aspects summarise these relationships: urban space utilisation, urban spatial form, and urban transportation infrastructure (Tian et al. 2020). Urban density (population or employment density) and land-use mix define the urban space utilisation. Increases in urban density and mixed function can effectively reduce per capita car use by reducing the number of trips and shortening travel distances. Similarly, the continuity of urban space and the dispersion of centres reduces travel distances (Tian et al. 2020), though such changes are rarely achieved without shifting transport infrastructure investments away from road capacity increases (Newman and Kenworthy 2015; McIntosh et al. 2017). For example, increased investment in public transport coverage, optimal transfer plans, shorter transit travel time, and improved transit travel efficiency make public transit more attractive (Heinen et al. 2017; Nugroho et al. 2018a; Nugroho et al. 2018b) and hence increase density and land values (Sharma and Newman 2020). Similarly, forgoing the development of major roads for the development of pedestrian and bike pathways enhances the attractiveness of active transport modes (Zahabi et al. 2016; Keall et al. 2018; Tian et al. 2020).

Ultimately, infrastructure investments influence the structural dependence on cars, which in turn influence the lock-in or path dependency of transport options with their greenhouse emissions (Newman et al. 2015; Grieco and Urry 2016). The 21st century saw a new trend to reach peak car use in some countries as a result of a revival in walking and transit use(Grieco and Urry 2016; Newman et al. 2017; Gota et al. 2019). While some cities continue on a trend towards reaching peak car use on a per-capita basis, for example Shanghai and Beijing (Gao and Newman 2020), there is a need for increased investments in urban form strategies that can continue to reduce car dependency around the world.

Behaviour continues to be a major source of interest in the decarbonisation of transport as it directly addresses demand. Behaviour is about people’s actions based on their preferences. Chapter 5 described an ‘Avoid, Shift, Improve’ process for demand-side changes that affect sectoral emissions. This section discusses some of the drivers of behaviour related to the transport sector and how they link to this ‘Avoid, Shift, Improve’ process.

Avoid: the effect of prices and income on demand. Research has shown that household income and price have a strong influence on people’s preferences for transport services (Bakhat et al. 2017; Palmer et al. 2018). The relationship between income and demand is defined by the income elasticity of demand. For example, research suggests that in China, older and wealthier populations continued to show a preference for car travel (Yang et al. 2019) while younger and low-income travellers sought variety in transport modes (Song et al. 2018). Similarly, Bergantino et al. (2018b) evaluated the income elasticity of transport by mode in the UK. They found that the income elasticity for private cars is 0.714, while the income elasticities of rail and bus use are 3.253 (the greater elasticity, the more the demand will grow or decline, depending on income). Research has also shown a positive relationship between income and demand for air travel, with income elasticities of air travel demand being positive and as large as 2 (Gallet and Doucouliagos 2014; Valdes 2015; Hakim and Merkert 2016; Hakim and Merkert 2019; Hanson et al. 2022). A survey in 98 Indian cities also showed income as the main factor influencing travel demand (Ahmad and de Oliveira 2016). Thus, as incomes and wealth across the globe rise, demand for travel is likely to increase as well.

The price elasticity of demand measures changes in demand as a result of changes in the prices of the services. In a meta-analysis of the price elasticity of energy demand, Labandeira et al. (2017) report the average long-term price elasticity of demand for gasoline and diesel to be –0.773 and –0.443, respectively. That is, demand will decline with increasing prices. A similar analysis of long-term data in the United States (US), the United Kingdom (UK), Sweden, Australia, and Germany reports the gasoline price elasticity of demand for car travel (as measured through vehicle-kilometre – vkm – per capita) ranges between –0.1 and –0.4 (Bastian et al. 2016). For rail travel, the price elasticity of demand has been found to range between –1.05 and –1.1 (Zeng et al. 2021). Similarly, price elasticities for air travel range from –0.53 to –1.91 depending on various factors such as purpose of travel (business or leisure), season, and month and day of departure (Morlotti et al. 2017). The price elasticities of demand suggest that car use is inelastic to prices, while train use is relatively inelastic to the cost of using rail. Conversely, consumers seem to be more responsive to the cost of flying, so that strategies that increase the cost of flying are likely to contribute to some avoidance of aviation-related GHG emissions.

While the literature continues to show that time, cost, and income dominate people’s travel choices (Ahmad and de Oliveira 2016; Capurso et al. 2019; He et al. 2020), there is also evidence of a role for personal values, and environmental values in particular, shaping choices within these structural limitations (Bouman and Steg 2019). For example, individuals are more likely to drive less when they care about the environment (De Groot et al. 2008; Abrahamse et al. 2009; Jakovcevic and Steg 2013; Hiratsuka et al. 2018; Ünal et al. 2019). Moreover, emotional and symbolic factors affect the level of car use (Steg 2005). Differences in behaviour may also result due to differences in gender, age, norms, values, and social status. For example, women have been shown to be more sensitive to parking pricing than men (Simićević et al. 2020).

Finally, structural shocks, such as a financial crisis, a pandemic, or the impacts of climate change could affect the price and income elasticities of demand for transport services (van Ruijven et al. 2019). COVID-19 lockdowns reduced travel demand by 19% (aviation by 32%) and some of the patterns that have emerged from the lockdowns could permanently change the elasticity of demand for transport (Tirachini and Cats 2020; Hendrickson and Rilett 2020; Newman 2020a; SLoCaT 2021; Hanson et al. 2022). In particular, the COVID-19 lockdowns have spurred two major trends: electronic communications replacing many work and personal travel requirements; and revitalised local active transport and e-micromobility (Newman 2020a; SLoCaT 2021). The permanence of these changes post-COVID-19 is uncertain but possible (Earley and Newman 2021) (Cross-Chapter Box 1 in Chapter 1). However, these changes will require growth of infrastructure for better ICT bandwidths in developing countries, and better provision for micromobility in all cities.

Shift: mode choice for urban and intercity transport. Shifting demand patterns (as opposed to avoiding demand) can be particularly important in decarbonising the transport sector. As a result, the cross-elasticity of demand across transport modes is of particular interest for understanding the opportunities for modal shift. The cross-elasticity represents the demand effect on mode i (e.g., bus) when an attribute of mode j (e.g., rail) changes marginally. Studies on the cross-elasticities of mode choice for urban travel suggest that the cross-elasticity for car demand is low, but the cross-elasticities of walking, bus, and rail with respect to cars are relatively large (Fearnley et al. 2017; Wardman et al. 2018). In practice, these cross-elasticities suggest that car drivers are not very responsive to increased prices for public transit, but transit users are responsive to reductions in the cost of driving. When looking at the cross-elasticities of public transit options (bus vs metro vs rail), research suggests that consumers are particularly sensitive to in-vehicle and waiting time when choosing public transit modes (Fearnley et al. 2018). These general results provide additional evidence that increasing the use of active and public transport requires interventions that make car use more expensive while making public transit more convenient (e.g., smart apps that tell the user the exact time for transit arrival (Box 10.1)).

The literature on mode competition for intercity travel reveals that while cost of travel is a significant factor (Zhang et al. 2017), sensitivity decreases with increasing income as well as when the cost of the trip was paid by someone else (Capurso et al. 2019). Some research suggests little competition between bus and air travel but the cross-elasticity between air and rail suggest strong interactions (Wardman et al. 2018). Price reduction strategies such as discounted rail fares could enhance the switch from air travel to high-speed rail. Both air fares and flight frequency impact high-speed rail (HSR) usage (Zhang et al. 2019b). Airline companies reduce fares on routes that are directly competing with HSR (Bergantino et al. 2018a) and charge high fares on non-HSR routes (Xia and Zhang 2016). On the Rome to Milan route, better frequency and connections, and low costs of HSR resulting from competition between HSR companies have significantly reduced air travel and shares of buses and cars (Desmaris and Croccolo 2018).

Finally, and as noted in Chapter 5, recent research shows that individual, social, and infrastructure factors also affect people’s mode choices. For example, perceptions about common travel behaviour (what people perceive to be ‘normal’ behaviour) influence their travel mode choice. The research suggests that well-informed individuals whose personal norms match low-carbon objectives, and who believe they have control over their decisions, are most motivated to shift mode. Nonetheless, such individual and social norms can only marginally influence mode choice unless infrastructure factors can enable reasonable time and cost savings (Convery and Williams 2019; Javaid et al. 2020; Feng et al. 2020; Wang et al. 2021).

Improve: consumer preferences for improved and alternative vehicles. While reductions in demand for travel and changes in mode choice can contribute to reducing GHG emissions from the transport sector, cars are likely to continue to play a prominent role. As a result, improving the performance of cars will be crucial for the decarbonisation of the transport sector. Sections 10.3 and 10.4 describe the technological options available for reduced CO2 emissions from vehicles. The effectiveness in deploying such technologies will partly depend on consumer preferences and their effect on adoption rates. Given the expanded availability of electric vehicles, there is also a growing body of work on the drivers of vehicle choice. A survey in Nanjing found women had more diverse travel purposes than men, resulting in a greater acceptance of electric bikes (Lin et al. 2017). Individuals are more likely to adopt an electric vehicle (EV) when they think this adoption benefits the environment or implies a positive personal attribute (Noppers et al. 2014; Noppers et al. 2015; Haustein and Jensen 2018). Other work suggests that people’s preference for EVs depends upon vehicle attributes, infrastructure availability, and policies that promote EV adoption, specifically, purchasing and operating costs, driving range, charging duration, vehicle performance, and brand diversity (Liao et al. 2016). Behaviour change to enable transport transformations will need to make the most of these factors while also working on the more structural issues of time, space, and cost.

Structural and behavioural choices that drive transport-related GHG emissions, such as time and cost based on geography of freight and urban fabric, are likely to continue to be major factors. But there is also a variation within each structural choice that is based around personal demand factors related to values that indirectly change choices in transport. Chapter 5 identified three megatrends that affect demand for services, including circular economy, the shared economy, and digitalisation. These three megatrends can have specific effect on transport emissions, as described below.

Circular economy. The problem of resources and their environmental impacts is driving the move to a circular economy (Bleischwitz et al. 2017). Circular economy principles include increased material efficiency, reusing or extending product lifetimes, recycling, and green logistics. Dematerialisation, the reduction in the quantity of the materials used in the production of one unit of output, is a circular economy principle that can affect the operations and emissions of the transport sector, as reductions in the quantities of materials used reduce transport needs, while reductions in the weight of products improve the efficiency of transporting them. Dematerialisation can occur through more efficient production processes but also when a new product is developed to provide the same functionality as multiple products. The best example of this trend is a smart phone, which provides the service of at least 22 other former devices (Rifkin 2019). A move to declutter lifestyles can also drive dematerialisation (Whitmarsh et al. 2017). Some potential for dematerialisation has been suggested due to 3-D printing, which would also reduce transport emissions through localised production of product components (d’Aveni 2015; UNCTAD 2018). There is evidence to suggest, however, that reductions in material use resulting from more efficient product design or manufacturing are offset by increased consumer demand (Kasulaitis et al. 2019). Whether or not dematerialisation can lead to reduction of emissions from the transport sector is still an open question that requires evaluating the entire product ecosystem (Van Loon et al. 2014; Coroama et al. 2015; Kasulaitis et al. 2019).

Shared economy. Shared mobility is arguably the most rapidly growing and evolving sector of the sharing economy and includes bike sharing, e-scooter sharing, car sharing, and on-demand mobility (Greenblatt and Shaheen 2015). The values of creating a more shared economy are related to both reduced demand and greater efficiency, as well as the notion of community well-being associated with the act of sharing instead of simply owning for oneself (Maginn et al. 2018; Sharp 2018). The literature on shared mobility is expanding, but there is much uncertainty about the effect shared mobility will have on transport demand and associated emissions (Nijland and Jordy 2017; ITF 2018a; Tikoudis et al. 2021).

Asia represents the largest car-sharing region with 58% of worldwide membership and 43% of global fleets deployed (Dhar et al. 2020). Europe accounts for 29% of worldwide members and 37% of shared vehicle fleets (Shaheen et al. 2018). Ride-sourcing and carpooling systems are among the many new entrants in the short-term shared mobility options. On-demand transport options complemented with technology have enhanced the possibility of upscaling (Alonso-González et al. 2018). Car sharing could provide the same level of service as taxis, but taxis could be three times more expensive (Cuevas et al. 2016). The sharing economy, as an emerging economic-technological phenomenon (Kaplan and Haenlein 2010), is likely to be a key driver of demand for transport of goods although data shows increasing container movement due to online shopping (Suel and Polak 2018).

There is growing evidence that this more structured form of behavioural change through shared economy practices, supported by a larger group than a single family, has a much greater potential to save transport emissions, especially when complemented with decarbonised grid electricity (Greenblatt and Shaheen 2015; Sharp 2018). Carpooling, for example, could result in an 11% reduction in vehicle-kilometres and a 12% reduction in emissions, as carpooling requires less empty or non-productive passenger-kilometres (pkm) (ITF 2020a; ITF 2020b). However, the use of local shared mobility systems such as on-demand transport may create more transport emissions if there is an overall modal shift out of transit (ITF 2018a; Schaller 2018). Similarly, some work suggests that commercial shared vehicle services such as Uber and Lyft are leading to increased vehicle km travelled (and associated GHG emissions) in part due to deadheading (Schaller 2018; Tirachini and Gomez-Lobo 2020; Ward et al. 2021). Successful providers compete by optimising personal comfort and convenience rather than enabling a sharing culture (Eckhardt and Bardhi 2015), and concerns have been raised regarding the wider societal impacts of these systems and for specific user groups such as older people (Fitt 2018; Marsden 2018). Concerns have also been expressed over the financial viability of demand-responsive transport systems (Ryley et al. 2014; Marsden 2018), how the mainstreaming of shared mobility systems can be institutionalised equitably, and the operation and governance of existing systems that are only mode- and operator-focused (Akyelken et al. 2018; Jittrapirom et al. 2018; Pangbourne et al. 2020; Marsden 2018).

Digitalisation. In the context of the transport sector, digitalisation has enabled teleworking, which in turn reduces travel demand. On the other hand, the prevalence of online shopping, enabled by the digital economy, could have mixed effects on transport emissions (Le et al. 2021). For example, online shopping could reduce vehicle-kilometres travelled but the move to expedited or rush delivery could mitigate some benefits as it prevents consolidation of freight (Jaller and Pahwa 2020).

Digitalisation could also lead to systemic changes by enabling smart mobility. The smart mobility paradigm refers to the process and practices of assimilation of ICTs and other sophisticated high-technology innovations into transport (Noy and Givoni 2018). Smart mobility can be used to influence transport demand and efficiency (Benevolo et al. 2016). The synergies of emerging technologies (ICT, internet of things, big data) and shared economy could overcome some of the challenges facing the adoption of emerging technologies (Marletto 2014; Chen et al. 2016; Weiss et al. 2018; Taiebat and Xu 2019) and enable the expected large growth in emerging cities to be more sustainable (Docherty et al. 2018). However, ICT, in particular the internet of things (IoT), could also cause more global energy demand (Hittinger and Jaramillo 2019). Box 10.1 summarises the main smart technologies being adopted rapidly by cities across the world and their use in transport. There is a growing body of literature about the effect of smart technology (including sensors guiding vehicles) on the demand for transport services. Smart technologies can improve competitiveness of transit and active transport over personal vehicle use by combining the introduction of new electro-mobility that improves time and cost along with behaviour change factors (Pålsson et al. 2017; SLoCaT 2018a; SLoCaT 2018b; SLoCaT2021). However, it is unclear what the net effect of smart technology on GHG emissions from the transport sector will be (Debnath et al. 2014; Lenz and Heinrichs 2017).

Autonomous vehicles are the other emerging transport technology that have the potential to significantly improve ride quality and safety. Planes and high-speed trains are already largely autonomous as they are guided in all their movements, especially coming into stations and airports, although that does not necessarily mean they are driverless. Automation is also being used in new on-road transit systems like trackless trams (Ndlovu and Newman 2020)). Private vehicles are being fitted with more and more levels of autonomy and many are being trialled as ‘driverless’ in cities (Aria et al. 2016; Skeete 2018). If autonomous systems can be used to help on-road transit become more time- and cost-competitive with cars, then the kind of transformative and disruptive changes needed to assist decarbonisation of transport become more feasible (Bösch et al. 2018; Kassens-Noor et al. 2020; Abe 2021). Similarly, vehicle automation could improve vehicle efficiency and reduce congestion, which would in turn reduce emissions (Vahidi and Sciarretta 2018; Massar et al. 2021). On the other hand, if autonomous cars make driving more convenient, they could reduce demand for transit (Auld et al. 2017; Sonnleitner et al. 2021). Paradoxically, autonomous cars could provide access to marginal groups such as the elderly, people with disabilities, and those who cannot drive, which could in turn increase travel demand (as measured by pkm) (Harper et al. 2016).

Heavy haulage trucks in the mining industry are already autonomous (Gaber et al. 2021) and automation of long-haul trucks may happen sooner than automation of LDVs (Hancock et al. 2019). Autonomous trucks may facilitate route and speed optimisation, and reduce fuel use, which can in turn reduce emissions (Nasri et al. 2018; Paddeu and Denby 2021). There is growing interest in using drones for package delivery. Drones could have lower impacts than ground-based delivery and, if deployed carefully, drones could reduce energy use and GHG emissions from freight transport (Stolaroff et al. 2018). Overall, some commentators are optimistic that smart and autonomous technologies can transform the GHG emissions from the transport sector (Seba 2014; Rifkin 2019; Sedlmeir et al. 2020). Others are more sanguine unless policy interventions can enable the technologies to be used for purposes that include zero carbon and the SDGs (Faisal et al. 2019; Hancock et al. 2019).

The interactions between systemic factors set out here and technology factors discussed in much more detail in the next sections show that there is always going to be a need to integrate both approaches. Good technology that has the potential to transform transport will not be used unless it fulfils broad mobility and accessibility objectives related to time, cost, and well-being. Chapter 5 has set out three transport transformations based on demand-side factors with highly transformative potential. Table 10.3 provides a summary of these systemic changes and their likely impact on GHG emissions. Note that the quantitative estimates provided in the table may not be additive and the combined effect of these strategies on GHG emissions from the transport sector require additional analysis.

The average fuel consumption of new internal combustion engine (ICE) vehicles has improved significantly in recent years due to more stringent emissions regulations. However, improvements are now slowing down. The average fuel consumption of LDVs decreased by only 0.7 % between 2016 and 2017, reaching 7.2 litres of gasoline-equivalent (Lg-eq) per 100 km in 2017, much slower than the 1.8 % improvement per year between 2005 and 2016 (GFEI 2020). Table 10.4 summarises recent and forthcoming improvements to ICE technologies and their effect on emissions from these vehicles. However, these improvements are not sufficient to meet deep decarbonisation levels in the transport sector. While there is significant and growing interest in electric and fuel-cell vehicles, future scenarios indicate that a large number of LDV may continue to be operated by ICE in conventional, hybrid, and plug-in hybrid configurations over the next 30 years (IEA 2019a), unless they are regulated away through ICE vehicle sales bans (as some nations have announced) (IEA 2021a). Moreover, ICE technologies are likely to remain the prevalent options for shipping and aviation. Thus, reducing CO2 and other emissions from ICEs through the use of low-carbon or zero-carbon fuels is essential to a balanced strategy for limiting atmospheric pollutant levels. Such alternative fuels for ICE vehicles include natural gas-based fuels, biofuels, ammonia, and other synthetic fuels.

Table 10.4 | Engine technologies to reduce emissions from light-duty ICE vehicles and their implementation stage. Table nomenclature: GDI = Gasoline direct injection, VVT = Variable valve technology, CDA = Cylinder deactivation, CR = compression ratio, GDCI = Gasoline direct injection compression ignition, EGR = exhaust gas recirculation, RCCI = Reactivity controlled compression ignition, GCI = Gasoline compression ignition. Source: Joshi (2020).

Natural Gas. Natural gas could be used as an alternative fuel to replace gasoline and diesel. Natural gas in vehicles can be used as compressed natural gas (CNG) and liquefied natural gas (LNG). CNG is gaseous at relatively high pressure (10 to 25 megapascal (MPa)) and temperature (–40 to 30°C). In contrast, LNG is used in liquid form at relatively low pressure (0.1 MPa) and temperature (–160°C). Therefore, CNG is particularly suitable for commercial vehicles and light- to medium-duty vehicles, whereas LNG is better suited to replace diesel in HDVs (Dubov et al. 2020; Dziewiatkowski et al. 2020; Yaïci and Ribberink 2021). CNG vehicles have been widely deployed in some regions, particularly in Asian-Pacific countries. For example, there are about 6 million CNG vehicles in China, the most of any country (Qin et al. 2020). However, only 20% of vehicles that operate using CNG were originally designed as CNG vehicles, with the rest being gasoline-fuelled vehicles that have been converted to operate with CNG (Chala et al. 2018).

Natural gas-based vehicles have certain advantages over conventional fuel-powered ICE vehicles, including lower emissions of criteria air pollutants, no soot or particulate, low carbon to Hydrogen ratio, moderate noise, a wide range of flammability limits, and high octane numbers (Kim 2019; Bayat and Ghazikhani 2020). Furthermore, the technology readiness level (TRL) of natural gas vehicles is very high (TRL 8–9), with direct modification of existing gasoline and diesel vehicles possible (Transport and Environment 2018; Peters et al. 2021; Sahoo and Srivastava 2021). On the other hand, methane emissions from the natural gas supply chain and tailpipe CO2 emissions remain a significant concern (Trivedi et al. 2020). As a result, natural gas as a transition transportation fuel may be limited due to better alternative options being available and due to regulatory pressure to decarbonise the transport sector rapidly. For example, the International Maritime Office (IMO) has set a target of 40% less carbon intensity in shipping by 2030, which cannot be obtained by simply switching to natural gas.

Biofuels. Since AR5, the faster than anticipated adoption of electromobility, primarily for LDVs, has partially shifted the debate around the primary use of biofuels from land transport to the shipping and aviation sectors (IEA 2017a; Davis et al. 2018). At the same time, other studies highlight that biofuels may have to complement electromobility in road transport, particularly in developing countries, offering relevant mitigation opportunities in the short- and mid-term (up to 2050) (IEA 2021b). An important advantage of biofuels is that they can be converted into energy carriers compatible with existing technologies, including current powertrains and fuel infrastructure. Also, biofuels can diversify the supply of transport fuel, raise energy self-sufficiency in many countries, and be used as a strategy to diversify and strengthen the agro-industrial sector (Puricelli et al. 2021). The use of biofuels as a mitigation strategy is driven by a combination of factors, including not only the costs and technology readiness levels of the different biofuel conversion technologies, but also the availability and costs of both biomass feedstocks and alternative mitigation options, and the relative speed and scale of the energy transition in energy and transport sectors (Box 10.2).

Many studies have addressed the lifecycle emissions of biofuel conversion pathways for land transport, aviation, and marine applications (Koeble et al. 2017; Staples et al. 2018; Tanzer et al. 2019). Bioenergy technologies generally struggle to compete with existing fossil fuel-based ones because of the higher costs involved. However, the extent of the cost gap depends critically on the availability and costs of biomass feedstock (IEA 2021b). Ethanol from corn and sugarcane is commercially available in countries such as Brazil and the US. Biodiesel from oil crops and hydro-processed esters and fatty acids are available in various countries, notably in Europe and parts of Southeast Asia. On the infrastructure side, biomethane blending is being implemented in some regions of the US and Europe, particularly in Germany, with the help of policy measures (IEA 2021b). While many of these biofuel conversion technologies could also be implemented using seaweed feedstock options, these value chains are not yet mature (Jiang et al. 2016).

Technologies to produce advanced biofuels from lignocellulosic feedstocks have suffered from slow technology development and are still struggling to achieve full commercial scale. Their uptake is likely to require carbon pricing and/or other regulatory measures, such as clean fuel standards in the transport sector or blending mandates. Several commercial-scale advanced biofuels projects are in development in many parts of the world, encompassing a wide selection of technologies and feedstock choices, including carbon capture and sequestration (CCS) that supports carbon dioxide removal. The success of these projects is vital to moving forward the development of advanced biofuels and bringing many of the advanced biofuels value chains closer to the market (IEA 2021b). Finally, biofuel production and distribution supply chains involve notable transport and logistical challenges that need to be overcome (Mawhood et al. 2016; Skeer et al. 2016; IEA 2017a; Puricelli et al. 2021).

Table 10.5 summarises performance data for different biofuel technologies, while Figure 10.3 shows the technology readiness levels.

Table 10.5 | Ranges of efficiency, GHG emissions, and relative costs of selected biofuel conversion technologies for road, marine, and aviation biofuels.

Notes: aCalculated as liquid fuels output divided by energy in feedstock entering the conversion plant; bGHG emissions here refers only to the conversion process. Impacts form the different biomass options are not included here as they are addressed in Chapter 7; cOlofsson et al. (2017); dKoeble et al. (2017); eSimell et al. (2014); f de Souza Dias et al. (2015); gCastanheira et al. (2015); hTanzer et al. (2019); iKlein et al. (2018); jNarula et al. (2017); kde Jong et al. (2017); lSalman et al. (2017); mMoreira et al. (2014); Roy et al. (2015); Handler et al. (2016); nSalman et al. (2017); Moreira et al. (2014); Roy et al. (2015); Handler et al. (2016).

Within the aviation sector, jet fuels produced from biomass resources (so-called sustainable aviation fuels, or SAF) could offer significant climate mitigation opportunities under the right policy circumstances. Despite the growing interest in aviation biofuels, demand and production volumes remain negligible compared to conventional fossil aviation fuels. Nearly all flights powered by biofuels have used fuels derived from vegetable oils and fats, and the blending level of biofuels into conventional aviation fuels for testing is up to 50% today (Mawhood et al. 2016). To date, only one facility in the US is regularly producing sustainable aviation fuels based on waste oil feedstocks. The potential to scale up bio-based SAF volumes is severely restricted by the lack of low-cost and sustainable feedstock options (Chapter 7). Lignocellulosic feedstocks are considered to have great potential for the production of financially competitive bio-based SAF in many regions. However, production facilities involve significant capital investment and estimated levelised costs are typically more than twice the selling price of conventional jet fuel. In some cases (notably for vegetable oils), the feedstock price is already higher than that of fossil jet fuel (Mawhood et al. 2016). Some promising technological routes for producing SAF from lignocellulosic feedstocks are below technology readiness level (TRL) 6 (pilot scale), with just a few players involved in the development of these technologies. Although it would be physically possible to address the mid-century projections for substantial use of biofuels in the aviation sector (according to the International Energy Agency (IEA) and other sectoral organisations (ICAO 2017)), this fuel deployment scale could only be achieved with very large capital investments in bio-based SAF production infrastructure, and substantial policy support.

In comparison to the aviation sector, the prospects for technology deployment are better in the shipping sector. The advantage of shipping fuels is that marine engines have a much higher operational flexibility on a mix of fuels, and shipping fuels do not need to undergo as extensive refining processes as road and aviation fuels to be considered drop-in. However, biofuels in marine engines have only been tested at an experimental or demonstration stage, leaving open the question about the scalability of the operations, including logistics issues. Similar to the aviation sector, securing a reliable, sustainable biomass feedstock supply and mature processing technologies to produce price-competitive biofuels at a large scale remains a challenge for the shipping sector (Hsieh and Felby 2017). Other drawbacks include industry concerns about oxidation, storage, and microbial stability for less purified or more crude biofuels. Assuming that biofuels are technically developed and available for the shipping sector in large quantities, a wider initial introduction of biofuels in the sector is likely to depend upon increased environmental regulation of particulate and GHG emissions. Biofuels may also offer a significant advantage in meeting ambitious sulphur emission reduction targets set by the sectoral organisations. More extensive use of marine biofuels will most likely be first implemented in inner-city waterways, inland river freight routes, and coastal green zones. Given the high efficiency of the diesel engine, a large-scale switch to a different standard marine propulsion method in the near to medium-term future seems unlikely . Thus, much of the effort has been placed on developing biofuels compatible with diesel engines. So far, biodiesel blends look promising, as it is used in land transport. Hydrotreated vegetable oil (HVO) is also a technically good alternative and is compatible with current engines and supply chains, while the introduction of multifuel engines may open the market for ethanol fuels (Hsieh and Felby 2017).

Ammonia. At room temperature and atmospheric pressure, ammonia is a colourless gas with a distinct odour. Due to relatively mild conditions for liquefaction, ammonia is transferred and stored as a liquefied or compressed gas and has been used as an essential industrial chemical resource for many products. In addition, since ammonia does not contain carbon, it has attracted attention as a carbon-neutral fuel that can also improve combustion efficiency (Gill et al. 2012). Furthermore, ammonia could also serve as a hydrogen carrier and be used in fuel cells. These characteristics have driven increased interest in the low-carbon production of ammonia, which would have to be coupled to low-carbon hydrogen production (with low-carbon electricity providing the needed energy or with CCS).

For conventional internal combustion engines, the use of ammonia remains challenging due to the relatively low burning velocity and high ignition temperature. Therefore, Frigo and Gentili (2014) have suggested a dual-fuelled spark ignition engine operated by liquid ammonia and hydrogen, where hydrogen is generated from ammonia using the thermal energy of exhaust gas. On the other hand, the high-octane number of ammonia means good knocking resistance of spark ignition engines and is promising for improving thermal efficiency. For compression ignition engines, the high-ignition temperature of ammonia requires a high compression ratio, causing an increase in mechanical friction. Since Gray et al. (1966), many studies have shown that the compression ratio can be reduced by mixing ammonia with secondary fuels such as diesel and hydrogen with low self-ignition temperatures, as summarised by Dimitriou and Javaid (2020). Using a secondary fuel with a high cetane number and the adoption of a suitable fuel injection timing has enabled highly efficient combustion of compression ignition engines in the dual fuel mode with ammonia ratios up to 95% (Dimitriou and Javaid 2020). One major challenge for realising an ammonia-fuelled engine is the reduction of unburned ammonia, as described in Section 6.4.5 (Reiter and Kong 2011). Processes being examined include the use of exhaust gas recirculation (EGR) (Pochet et al. 2017) and after treatment systems. However, these processes require space, which is a constraint for LDVs and air transport but more practical for ships. Shipbuilders are developing an ammonia engine based on the existing diesel dual-fuel engine to launch a service in 2025 (Brown 2019; MAN-ES 2019). Ammonia could therefore contribute significantly to decarbonisation in the shipping sector (Section 10.6), with potential niche applications elsewhere.

Synthetic fuels. Synthetic fuels can contribute to transport decarbonisation through synthesis from electrolytic hydrogen produced with low-carbon electricity or hydrogen produced with CCS, and captured CO2 using the Fischer-Tropsch process (Liu et al. 2020a). Due to similar properties of synthetic fuels to those of fossil fuels, synthetic fuels can reduce GHG emissions in both existing and new vehicles without significant changes to the engine design. While the Fischer-Tropsch process is a well-established technology (Liu et al. 2020a), low-carbon synthetic fuel production is still at the demonstration stage. Even though their production costs are expected to decline in the future due to lower renewable electricity prices, increased scale of production, and learning effects, synthetic fuels are still up to three times more expensive than conventional fossil fuels (Section 6.6.2.4). Furthermore, since the production of synthetic fuels involves thermodynamic conversion loss, there is a concern that the total energy efficiency is lower than that of electric vehicles (Yugo and Soler 2019). Given these high costs and limited scales, the adoption of synthetic fuels will likely focus on the aviation, shipping, and long-distance road transport segments, where decarbonisation by electrification is more challenging. In particular, synthetic fuels are considered promising as an aviation fuel (Section 10.5).

Widespread electrification of the transport sector is likely crucial for reducing transport emissions and depends on appropriate electrical energy storage systems (EES). However, large-scale diffusion of EES depends on improvements in energy density (energy stored per unit volume), specific energy (energy stored per unit weight), and costs (Cano et al. 2018). Recent trends suggest EES-enabled vehicles are on a path to becoming the leading technology for LDVs, but their contribution to heavy-duty freight is more uncertain.

Electrochemical storage of light and medium-duty vehicles. Electrochemical storage, i.e., batteries, are one of the most promising forms of energy storage for the transport sector and have dramatically improved in their commerciality since AR5. Rechargeable batteries are of primary interest for applications within the transport sector, with a range of mature and emerging chemistries able to support the electrification of vehicles. The most significant change since AR5 and SPR1.5 is the dramatic rise in lithium-ion batteries (LIB), which has enabled electromobility to become a major feature of decarbonisation.

Before the recent growth in market share of LIBs, lead-acid batteries, nickel batteries, high-temperature sodium batteries, and redox flow batteries were of particular interest for the transport sector (Placke et al. 2017). Due to their low costs, lead-acid batteries have been used in smaller automotive vehicles, e.g., e-scooters and e-rickshaws (Dhar et al. 2017). However, their application in electric vehicles will be limited due to their low specific energy (Andwari et al. 2017). Nickel-metal hydride (NiMH) batteries have a better energy density than lead-acid batteries and have been well optimised for regenerative braking (Cano et al. 2018). As a result, NiMH batteries were the battery of choice for hybrid electric vehicles (HEVs). Ni-Cadmium (NiCd) batteries have energy densities lower than NiMH batteries and cost around ten times more than lead-acid batteries (Table 6.5). For this reason, NiCd batteries do not have major prospects within automotive applications. There are also no examples of high-temperature sodium or redox flow batteries being used within automotive applications.

Commercial application of LIBs in automotive applications started around 2000 when the price of LIBs was more than USD1000 per kWh (Schmidt et al. 2017). By 2020, the battery manufacturing capacity for automotive applications was around 300 GWh per year (IEA 2021a). Furthermore, by 2020, the average battery pack cost had come down to USD137 per kWh, a reduction of 89% in real terms since 2010 (Henze 2020). Further improvements in specific energy, energy density (Nykvist et al. 2015; Placke et al. 2017) and battery service life (Liu et al. 2017) of LIBs are expected through additional design optimisation (Table 6.5). These advances are expected to lead to EVs with even longer driving ranges, further supporting the uptake of LIBs for transport applications (Cano et al. 2018). However, the performance of LIBs under freezing and high temperatures is a concern (Liu et al. 2017) for reliability. Auto manufacturers have some pre-heating systems for batteries to see that they perform well in very cold conditions (Wu et al. 2020).

For EVs sold in 2018, the material demand was about 11 kilotonnes (kt) of optimised lithium, 15 kt of cobalt, 11 kt of manganese, and 34 kt of nickel (IEA 2019a; IEA 2021a). IEA projections for 2030 in the EV 30@30 scenario show that the demand for these materials would increase by 30 times for lithium and around 25 times for cobalt. While there are efforts to move away from expensive materials such as cobalt (IEA 2019a; IEA 2021a), dependence on lithium will remain, which may be a cause of concern (Olivetti et al. 2017; You and Manthiram 2018). A more detailed discussion on resource constraints for lithium is provided in Box 10.6.

Externalities from resource extraction are another concern, though current volumes of lithium are much smaller than other metals (steel, aluminium). As a result, lithium was not even mentioned in UNEP’s global resource outlook (IRP 2019). Nonetheless, it is essential to manage demand and limit externalities since the demand for lithium is going to increase many times in the future. Reuse of LIBs used in EVs for stationary energy applications can help in reducing the demand for LIBs. However, the main challenges are the difficulty in accessing the information on the health of batteries to be recycled and technical problems in remanufacturing the batteries for their second life (Ahmadi et al. 2017). Recycling lithium from used batteries could be another possible supply source (Winslow et al. 2018). While further R&D is required for commercialisation (Ling et al., 2018), recent efforts at recycling LIBs are very encouraging (Ma et al. 2021). The standardisation of battery modules and packaging within and across vehicle platforms, increased focus on design for recyclability, and supportive regulation are important to enable higher recycling rates for LIBs (Harper et al. 2019).

Several next-generation battery chemistries are often referred to as post-LIBs (Placke et al. 2017). These chemistries include metal-sulphur, metal-air, metal-ion (besides Li), and all-solid-state batteries. The long development cycles of the automotive industry (Cano et al. 2018) and the advantages of LIBs in terms of energy density and cycle life (Table 6.5) mean that it is unlikely that post-LIB technologies will replace LIBs in the next decade. However, lithium-sulphur, lithium-air, and zinc-air have emerged as potential alternatives for LIBs. These emerging chemistries may also be used to supplement LIBs in dual-battery configurations, to extend the driving range at lower costs or with higher energy density (Cano et al. 2018). Lithium-sulphur (Li-S) batteries have a lithium metal anode with a higher theoretical capacity than lithium-ion anodes and much lower-cost sulphur cathodes relative to typical Li-ion insertion cathodes (Manthiram et al. 2014). As a result, Li-S batteries are much cheaper than LIB to manufacture and have a higher energy density (Table 6.5). Conversely, these batteries face challenges from sulphur cathodes, such as low conductivity of the sulphur and lithium sulphide phases, and the relatively high solubility of sulphur species in common lithium battery electrolytes, leading to low cycle life (Cano et al. 2018). Lithium-air batteries offer a further improvement in specific energy and energy density above Li–S batteries owing to their use of atmospheric oxygen as a cathode in place of sulphur. However, their demonstrated cycle life is much lower (Table 6.5). Lithium-air batteries also have low specific power. Therefore, lithium-air require an extra battery for practical applications (Cano et al. 2018). Finally, zinc–air batteries could more likely be used in future EVs because of their more advanced technology status and higher practically achievable energy density (Fu et al. 2017). Like Li-air batteries, their poor specific power and energy efficiency will probably prevent zinc-air batteries from being used as a primary energy source for EVs. Still, they could be promising when used in a dual-battery configuration (Cano et al. 2018).

The technological readiness of batteries is a crucial parameter in the advancement of EVs (Manzetti and Mariasiu 2015). Energy density, power density, cycle life, calendar life, and the cost per kWh are the pertinent parameters for comparing the technological readiness of various battery technologies (Manzetti and Mariasiu 2015; Andwari et al. 2017; Lajunen et al. 2018). Table 6.5 provides a summary of the values of these parameters for alternative battery technologies. LIBs comprehensively dominate the other battery types and are at a readiness level where they can be applied for land transport applications (cars, scooters, electrically-assisted cycles) and at battery pack costs below USD150 per kWh, making EVs cost-competitive with conventional vehicles (Nykvist et al. 2019). In 2020 the stock of battery electric LDVs had crossed the 10 million mark (IEA 2021a). Schmidt et al. (2017) project that the cost of a battery pack for LIBs will reach USD100 per kWh by 2030, but more recent trends show this could happen much earlier. For example, according to IEA, battery pack costs could be as low as USD80 per kWh by 2030 (IEA 2019a). In addition, there are clear trends that now vehicle manufacturers are offering vehicles with bigger batteries, greater driving ranges, higher top speeds, faster acceleration, and all size categories (Nykvist et al. 2019). In 2020 there were over 600,000 battery electric buses and over 31,000 battery electric trucks operating globally (IEA 2021a).

LIBs are not currently envisaged to be suitable for long-haul transport. However, several battery technologies are under development (Table 6.5), which could further enhance the competitiveness of EVs and expand their applicability to very short-haul aviation and ships, especially smaller vehicles. Li-S, Li-air, and Zn-air hold the highest potential for these segments (Cano et al. 2018). All three of these technologies rely on making use of relatively inexpensive elements, which can help bring down battery costs (Cano et al. 2018). The main challenge these technologies face is in terms of the cycle life. Out of the three, Li-S has already been used for applications in unmanned aerial vehicles (Fotouhi et al., 2017) due to relatively high specific energy (almost double the state of the art LIBs). However, even with low cycle life, Li-air and Zn-air hold good prospects for commercialisation as range extender batteries for long-range road transport and with vehicles that are typically used for city driving (Cano et al. 2018).

Alternative electricity storage technologies for heavy-duty transport. While LIBs described in the previous section are driving the electrification of LDVs, their application to railways, aviation, ships, and large vehicles faces challenges due to the higher power requirements of these applications. The use of a capacitor with a higher power density than LIBs could be suitable for the electrification of such vehicles. It is one of the solutions for regenerating large and instantaneous energy from regenerative brakes. Classical capacitors generally show more attractive characteristics in power density (8000–10,000 watts per kilogram (W/kg)) than batteries. However, the energy density is poor (1–4 watt-hours per kilogram (Wh/kg)) compared to batteries, and there is an issue of self-discharge (González et al. 2016; Poonam et al. 2019). To improve the energy density, electrochemical double layer capacitors (EDLCs; supercapacitor) and hybrid capacitors (10–24 Wh/kg, 900–9000 W/kg at the product level) such as Li-ion capacitors have been developed. The highest energy density of the LIC system (100–140 Wh/kg in the research stage) are approaching that of the Li-ion battery systems (80–240 Wh/kg in the product stage) (Naoi et al. 2012; Panja et al. 2020). Examples of effective use of capacitors include a 12-tonne truck with a capacitor-based kinetic energy recovery system that has been reported to save up to 32% of the fuel use of a standard truck (Kamdar 2017). Similarly, an EDLC bank applied to electric railway systems has been shown to result in a 10% reduction in power consumption per day (Takahashi et al. 2017). Finally, systems in which capacitors are mounted on an electric bus for charging at a stop have been put into practical use, for example by a trackless tram (Newman et al. 2019). At the bus stop, the capacitor is charged at 600 kW for 10 about 40 seconds, which provides enough power for about 5 to 10 km (Newman et al. 2019). In addition, more durable capacitors can achieve a longer life than LIB systems (ADB 2018).

Hybrid energy storage (HES) systems, which combine a capacitor and a battery, achieve both high power and high energy, solving problems such as capacity loss of the battery and self-discharge of the capacitor. In these systems, the capacitor absorbs the steeper power, while the LIB handles the steady power, thereby reducing the power loss of the EV to half. Furthermore, since the in-rush current of the battery is suppressed, there is an improvement in the reliability of the LIB (Noumi et al. 2014). In a hybrid diesel train, 8.2% of the regenerative energy is lost due to batteries’ limited charge-discharge performance; however, using an EDLC with batteries can save this energy (Takahashi et al. 2017; Mayrink et al. 2020).

The development of power storage devices and advanced integrated system approaches, including power electronics circuits such as HES and their control technologies, are important for the electrification of mobility. These technologies are solutions that could promote the electrification of systems, reduce costs, and contribute to the social environment through multiple outcomes in the decarbonisation agenda.

In harder-to-electrify transport segments, such as heavy-duty vehicles, shipping, and aviation, hydrogen holds significant promise for delivering emissions reductions if it is produced using low-carbon energy sources. In particular, hydrogen fuel cells are seen as an emerging option to power larger vehicles for land-based transport (Tokimatsu et al. 2016; IPCC 2018; IEA 2019b). Despite this potential, further advancements in technological and economic maturity will be required in order for hydrogen fuel cells to play a greater role. While this section focuses primarily on hydrogen fuel cells, ammonia and methanol fuel cells may also emerge as options for low power applications.

During the last decade, hydrogen fuel cell vehicles (HFCVs) have attracted growing attention, with fuel cell technology improving through research and development. Fuel cell systems cost 80% to 95% less than they did in the early 2000s, at approximately USD50 per kW for light-duty (80 kW) and $100 per kW for medium-heavy-duty (160 kW). These costs are approaching the US Department of Energy’s (US DOE) goal of USD40 per kW in 2025 at a production target of 500,000 systems per year (IEA 2019c). In addition to cost reductions, the power density of fuel cell stacks has now reached around 3.0 kilowatt per litre (kW/l) and average durability has improved to approximately 2000 to 3000 hours (Jouin et al. 2016; Kurtz et al. 2019). Despite these improvements, fuel cell systems are not yet mature for many commercial applications. For example, the US DOE has outlined that for hydrogen fuel cell articulated trucks (semi-trailers) to compete with diesel vehicles, fuel cell durability will need to reach 30,000 hours (US DOE 2019). While some fuel cell buses have demonstrated durability close to these targets (Eudy and Post 2018a), another review of light fuel cell vehicles found maximum durability of 4000 hours (Kurtz et al. 2019). As more fuel cell vehicles are trialled, it is expected that further real-world data will become available to track ongoing fuel cell durability improvements.

Ammonia and methanol fuel cells are considered to be less mature than hydrogen fuel cells. However, they offer the benefit of using a more easily transported fuel that can be directly used without converting to hydrogen (Zhao et al. 2019). Conversely, both methanol and ammonia are toxic, and in the case of methanol fuel cells, carbon dioxide is released as a by-product of generating electricity with the fuel cell (Zhao et al. 2019). Due to the lower power output, methanol and ammonia fuel cells are also not well suited to heavy-duty vehicles (Jeerh et al. 2021). They are therefore unlikely to compete with hydrogen fuel cells. However, ammonia and methanol could be converted to hydrogen at refuelling stations as an alternative to being directly used in fuel cells (Zhao et al. 2019).

Several FCV-related technologies are fully ready for demonstration and early market deployment, however, further research and development will be required to achieve full-scale commercialisation, likely from 2030 onwards (Staffell et al. 2019; Energy Transitions Commission 2020; IEA 2021b). Some reports argue that it may be possible to achieve serial production of fuel cell heavy-duty trucks in the late 2020s, with comparable costs to diesel vehicles achieved after 2030 (Jordbakker et al. 2018). Over the next decade or so, hydrogen FCVs could become cost-competitive for various transport applications, potentially including long-haul trucks, marine ships, and aviation (Hydrogen Council 2017; FCHEA 2019; FCHJU 2019; BloombergNEF 2020; Hydrogen Council 2020). The speed of fuel cell system cost reduction is a key factor for achieving widespread uptake. Yet, experts disagree on the relationship between the scale of fuel cell demand, cost, and performance improvements (Cano et al. 2018). Costs of light-, medium-, and heavy-duty fuel cell powertrains have decreased by orders of magnitude with further reductions of a factor of two expected with continued technological progress (Whiston et al. 2019). For example, the costs of platinum for fuel cell stacks have decreased by an order of magnitude (Staffell et al. 2019); current generation FCVs use approximately 0.25 g/kW platinum and a further reduction of 50–80% is expected by 2030 (Hao et al. 2019).

Hydrogen is likely to take diverse roles in the future energy system: as a fuel in industry and buildings, as well as transport, and as energy storage for variable renewable electricity. Further research is required to understand better how a hydrogen transport fuel supply system fits within the larger hydrogen energy system, especially in terms of integration within existing infrastructure, such as the electricity grid and the natural gas pipeline system (IEA 2015).

Strong and durable policies would be needed to enable widespread use of hydrogen as a transport fuel and to sustain momentum during a multi-decade transition period for hydrogen FCVs to become cost-competitive with electric vehicles (Hydrogen Council 2017; FCHEA 2019; FCHJU 2019; IEA 2019c; BNEF 2020; Hydrogen Council 2020). The analysis suggests that hydrogen is likely to have strategic and niche roles in transport, particularly in long-haul shipping and aviation. With continuing improvements, hydrogen and electrification will likely play a role in decarbonising heavy-duty road and rail vehicles.

The transport sector relies on liquid gasoline, and diesel for land-based transport, jet fuel for aviation, and heavy fuel oil for shipping. Extensive infrastructure for refuelling liquid fossil fuels already exists. Ammonia, synthetic fuels, and biofuels have emerged as alternative fuels for powering combustion engines and turbines used in land, shipping, and aviation (Figure 10.2). Synthetic fuels such as e-methanol and Fischer-Tropsch liquids have similar physical properties and could be used with existing fossil fuel infrastructure (Yugo and Soler, 2019). Similarly, biofuels have been used in several countries together with fossil fuels (Panoutsou et al. 2021). Ammonia is a liquid, but only under pressure, and therefore will not be compatible with liquid fossil fuel refuelling infrastructure. Ammonia is, however, widely used as a fertiliser and chemical raw material and 10% of annual ammonia production is transported via sea (Gallucci 2021). As such, a number of port facilities include ammonia storage and transport infrastructure and the shipping industry has experience in handling ammonia (Gallucci 2021). This infrastructure would likely need to be extended in order to support the use of ammonia as a fuel for shipping and therefore ports are likely to be the primary sites for these new refuelling facilities.

EVs and HFCV require separate infrastructure than liquid fuels. The successful diffusion of new vehicle technologies is dependent on the preceding deployment of infrastructure (Leibowicz 2018), so that the deployment of new charging and refuelling infrastructure will be critical for supporting the uptake of emerging transport technologies like EVs and HFCVs, where it makes sense for each to be deployed. As a result, there is likely a need for simultaneous investment in both infrastructure and vehicle technologies to accelerate decarbonisation of the transport sector.

Charging infrastructure. Charging infrastructure is important for a number of key reasons. From a consumer perspective, robust and reliable charging infrastructure networks are required to build confidence in the technology and overcome the often-cited barrier of ‘range anxiety’ (She et al. 2017). Range anxiety is where consumers do not have confidence that an EV will meet their driving range requirements. For LDVs, the majority of charging (75–90%) has been reported to take place at or near homes (Figenbaum 2017; Webb et al. 2019; Wenig et al. 2019). Charging at home is a particularly significant factor in the adoption of EVs as consumers are less willing to purchase an EV without home charging (Berkeley et al. 2017; Funke and Plötz 2017; Nicholas et al. 2017). However, home charging may not be an option for all consumers. For example, apartment dwellers may face specific challenges in installing charging infrastructure (Hall and Lutsey 2020). Thus, the provision of public charging infrastructure is another avenue for alleviating range anxiety, facilitating longer distance travel in EVs, and in turn, encouraging adoption (Hall and Lutsey 2017; Melliger et al. 2018; Narassimhan and Johnson 2018; Melton et al. 2020). Currently, approximately 10% of charging occurs at public locations, roughly split equally between alternating current (AC) (slower) and direct current (DC) (fast) charging (Figenbaum 2017; Webb et al. 2019; Wenig et al. 2019). Deploying charging infrastructure at workplaces and commuter car parks is also important, particularly as vehicles are parked at these locations for many hours. Indeed, around 15–30% of EV charging currently occurs at these locations (Figenbaum 2017; Webb et al. 2019; Wenig et al. 2019). It has been suggested that automakers and utilities could provide support for the installation of home charging infrastructure (Hardman et al. 2018), while policymakers can provide support for public charging. Such support could come via supportive planning policy, building regulations, and financial support. Policy support could also incentivise the deployment of charging stations at workplaces and commuter car parks. Charging at these locations would have the added benefit of using excess solar energy generated during the day (Hardman et al. 2018; Webb et al. 2019).

While charging infrastructure is of high importance for the electrification of light-duty vehicles, arguably it is even more important for heavy-duty vehicles, given the costs of high-power charging infrastructure. It is estimated that the installed cost of fast-charging hardware can vary between approximately USD45,000 to USD200,000 per charger, depending on the charging rate, the number of chargers per site, and other site conditions (Hall and Lutsey 2019; Nelder and Rogers 2019; Nicholas 2019). Deployment of shared charging infrastructure at key transport hubs, such as bus and truck depots, freight distribution centres, marine shipping ports and airports, can encourage a transition to electric vehicles across the heavy transport segments. Furthermore, if charging infrastructure sites are designed to cater for both light- and heavy-duty vehicles, infrastructure costs could decrease by increasing utilisation across multiple applications and/or fleets (Nelder and Rogers 2019).

There are two types of charging infrastructure for electric vehicles: conductive charging involving a physical connection and wireless/induction charging. The majority of charging infrastructure deployed today for light- and heavy-duty vehicles is conductive. However, wireless charging technologies are beginning to emerge – particularly for applications like bus rapid transit – with vehicles able to charge autonomously while parked and/or in motion (IRENA 2019). For road vehicles, electric road systems, or road electrification, is also emerging as an alternative form of conductive charging infrastructure that replaces a physical plug (Ainalis et al. 2020; Hill et al. 2020). This type of charging infrastructure is particularly relevant for road freight where load demand is higher. Road electrification can take the form of a charging rail built into the road pavement, run along the side of the road, through overhead catenary power lines – similar to electrical infrastructure used for rail – or at recharging facilities at stations along the route. This infrastructure can also be used to directly power other electrified powertrains, such as hybrid and HFCV (Hardman et al. 2018; Hill et al. 2020).

Charging infrastructure also varies in terms of the level of charging power. For light vehicles, charging infrastructure is generally up to 350 kW, which provides approximately 350 kilometres for every 10 minutes of charging. For larger vehicles, like buses and trucks, charging infrastructure is generally up to 600 kW, providing around 50–100 km for every 10 minutes of charging (depending on the size of the vehicle). Finally, even higher-power charging infrastructure is currently being developed at rates greater than 1 MW, particularly for long-haul trucks and for short-haul marine shipping and aviation. For example, one of the largest electric ferries in the world, currently operating in Denmark, uses a 4.4 MW charger (Heinemann et al. 2020).

Finally, there are several different charging standards, varying across transport segments and across geographical locations. Like electrical appliances, different EV charging connectors and sockets have emerged in different regions, such as CCS2 in Europe (ECA 2021), GB/T in China (Hove and Sandalow 2019). Achieving interoperability between charging stations is seen as another important issue for policymakers to address to provide transparent data to the market on where EV chargers are located and a consistent approach to paying for charging sessions (van der Kam and Bekkers 2020). Interoperability could also play an important role in enabling smart charging infrastructure (Neaimeh and Andersen 2020).

Smart charging: electric vehicle-grid integration strategies. EVs provide several opportunities for supporting electricity grids if appropriately integrated. Conversely, a lack of integration could negatively affect the grid, particularly if several vehicles are charged in parallel at higher charging rates during peak demand periods (Webb et al. 2019; Jochem et al. 2021). There are three primary approaches to EV charging. In unmanaged charging, EVs are charged ad hoc, whenever connected, regardless of conditions on the broader electricity grid (Webb et al. 2019; Jochem et al. 2021). Second, in managed charging, EVs are charged during periods beneficial to the grid, e.g., at periods of high renewable generation and/or low demand. Managed charging also allows utilities to regulate the rate of charge and can thus provide frequency and regulation services to the grid (Weis et al. 2014). Finally, in bidirectional charging or vehicle-to-grid (V2G), EVs are generally subject to managed charging, but an extension provides the ability to export electricity from the vehicle’s battery back to the building and/or wider electricity grid (Ercan et al. 2016; Noel et al. 2019; Jochem et al. 2021). The term ‘smart charging’ has become an umbrella term to encompass both managed charging (often referred to as V1G) and V2G. For electric utilities, smart charging strategies can provide back-up power, support load balancing, reduce peak loads (Zhuk et al. 2016; Noel et al. 2019; Jochem et al. 2021), reduce the uncertainty in forecasts of daily and hourly electrical loads (Peng et al. 2012), and allow greater utilisation of generation capacity (Hajimiragha et al. 2010; Madzharov et al. 2014).

Smart charging strategies can also enhance the climate benefits of EVs (Yuan et al. 2021). Controlled charging can help avoid high-carbon electricity sources, decarbonisation of the ancillary service markets, or peak shaving of high-carbon electricity sources (Jochem et al. 2021). V2G-capable EVs can result in even lower total emissions, particularly when compared to other alternatives (Reddy et al. 2016). Noel et al. (2019) analysed V2G pathways in Denmark and noted that at a penetration rate of 75% by 2030, USD34 billion in social benefits could be accrued (through things like displaced pollution). These social benefits translate to USD1,200 per vehicle. V2G-capable EVs were found to have the potential to reduce carbon emissions compared to a conventional gasoline vehicle by up to 59%, assuming optimised charging schedules (Hoehne and Chester 2016).

Projections of energy storage suggest smart charging strategies will come to play a significant role in future energy systems. Assessment of different energy storage technologies for Europe showed that V2G offered the most storage potential compared to other options and could account for 200 GW of installed capacity by 2060, whereas utility-scale batteries and pumped hydro storage could provide 160 GW of storage capacity (Després et al. 2017). Another study found that EVs with controlled charging could provide similar services to stationary storage but at a far lower cost (Coignard et al. 2018). While most deployments of smart charging strategies are still at the pilot stage, the number of projects continues to expand, with the V2G Hub documenting at least 90 V2G projects across 22 countries in 2021 (Vehicle to Grid 2021). Policymakers have an important role in facilitating collaboration between vehicle manufacturers, electricity utilities, infrastructure providers, and consumers to enable smart charging strategies and ensure EVs can support grid stability and the uptake of renewable energy. This is a critical part of decarbonising transport.

Hydrogen infrastructure. HFCVs are reliant on the development of widespread and convenient hydrogen refuelling stations (FCHEA 2019; IEA 2019c; BNEF 2020). Globally, there are around 540 hydrogen refuelling stations, with the majority located in North America, Europe, Japan, and China (IEA 2021a). Approximately 70% of these refuelling stations are open to the public (Coignard et al. 2018). Typical refuelling stations currently have a refuelling capacity of 100 to 350 kg/day (CARB 2019; CARB 2020; H2 Tools 2020; AFDC 2021). At most, current hydrogen refuelling stations have daily capacities under 500 kg a day (Liu et al. 2020b).

The design of hydrogen refuelling stations depends on the choice of methods for hydrogen supply and delivery, compression and storage, and the dispensing strategy. Hydrogen supply could happen via on-site production or via transport and delivery of hydrogen produced off-site. At the compression stage, hydrogen is compressed to achieve the pressure needed for economic stationary and vehicle storage. This pressure depends on the storage strategy. Hydrogen can be stored as a liquid or a gas. Hydrogen can also be dispensed to vehicles as a gas or a liquid, depending on the design of the vehicles (though it tests the extremes of temperature range and storage capacity for an industrial product). The technological and economic development of each of these components continues to be researched.

If hydrogen is produced off site in a large centralised plant, it must be stored and delivered to refuelling stations. The cost of hydrogen delivery depends on the amount of hydrogen delivered, the delivery distance, the storage method (compressed gas or cryogenic liquid), and the delivery mode (truck or pipeline). Table 10.6 describes the three primary options for hydrogen delivery. Most hydrogen refuelling stations today are supplied by trucks and, very occasionally, hydrogen pipelines. Gaseous tube trailers could also be used to deliver hydrogen in the near term, or over shorter distances, due to the low fixed cost (although the variable cost is high). Both liquefied truck trailers and pipelines are recognised as options in the medium to long term as they have higher capacities and lower costs over longer distances (FCHJU 2019; Li et al. 2020; EU 2021). Alternatively, hydrogen can be produced on site using a small-scale on-site electrolyser or steam methane reforming unit combined with CCS. Hydrogen is generally dispensed to vehicles as a compressed gas at pressures 350 or 700 bar, or as liquified hydrogen at –253°C (Hydrogen Council 2020).

Table 10.6 | Overview of three transport technologies for hydrogen delivery in the transport sector showing relative differences. Source: IEA (2019c).

The costs for hydrogen refuelling stations vary widely and remain uncertain for the future (IEA 2019c). The IEA reports that the investment cost for one hydrogen refuelling station ranges between USD0.6 million and USD2 million for hydrogen at a pressure of 700 bar and a delivery capacity of 1300 kg per day. The investment cost of hydrogen refuelling stations with lower refuelling capacities (~50 kg H2 per day) delivered at lower pressure (350 bar) range between USD0.15–1.6 million. A separate estimate by the International Council for Clean Transport suggests that at a capacity of 600 kg of hydrogen per day, the capital cost of a single refuelling station would be approximately USD1.8 million (ICCT 2017). Given the high investment costs for hydrogen refuelling stations, low utilisation can translate into a high price for delivered hydrogen. In Europe, most pumps operate at less than 10% capacity. For small refuelling stations with a capacity of 50 kg H2 per day, this utilisation rate translates to a high price of around USD15–25 per kg H2– in line with current retail prices (IEA 2019c). The dispensed cost of hydrogen is also highly correlated with the cost of electricity, when H2 is produced using electrolysis, which is required to produce low-carbon hydrogen.

LDVs represent the main mode of transport for private citizens (ITF 2019) and currently represent the largest share of transport emissions globally (IEA 2019d). Currently, powertrains depending on gasoline and diesel fuels remain the dominant technology in the LDV segment (IEA 2019d). HEVs, and fully battery electric vehicles (BEVs), however, have become increasingly popular in recent years (IEA 2021a). Correspondingly, the number of lifecycle assessment (LCA) studies investigating HEVs, BEVs, and fuel cell vehicles have increased. While historically the focus has been on the tailpipe emissions of LDVs, LCA studies demonstrate the importance of including emissions from the entire vehicle value chain, particularly for alternative powertrain technologies.

Figure 10.4 presents the cumulative lifecycle emissions for selected powertrain technologies and fuel chain combinations for compact and mid-sized LDVs. This figure summarises the harmonised findings from the academic literature reviewed and the data submitted through an IPCC data collection effort, as described in Appendix 10.1 (Hawkins et al. 2013; Messagie et al. 2014; Bauer et al. 2015; Tong et al. 2015b; Ellingsen et al. 2016; Gao et al. 2016; Kim and Wallington 2016; Cai et al. 2017; Evangelisti et al. 2017; Ke et al. 2017; Lombardi et al. 2017; Miotti et al. 2017; Valente et al. 2017; Cox et al. 2018; de Souza et al. 2018; Elgowainy et al. 2018; Luk et al. 2018; Bekel and Pauliuk 2019; Cusenza et al. 2019; Hoque et al. 2019; IEA 2019a; Rosenfeld et al. 2019; Shen et al. 2019; Wang et al. 2019; Wu et al. 2019; Ambrose et al. 2020; Benajes et al. 2020; Hill et al. 2020; Knobloch et al. 2020; Prussi et al. 2020; Qiao et al. 2020; Wolfram et al. 2020; Zheng et al. 2020; Sacchi 2021; Valente et al. 2021). The values in the figure (and the remaining figures in this section) depend on the 100-year global warming potential (GWP) used in each study, which may differ from the recent GWP updates from WGI. However, it is unlikely that the qualitative insights gained from the figures in this section would change using the update 100-year GWP values.

Furthermore, note that the carbon footprint of biofuels used in Figure 10.4 are aggregate numbers not specific to any individual value chain or fuel type. They are derived by combining land use-related carbon emissions from Chapter 7 with conversion efficiencies and emissions as described in Section 10.3. Specifically, land-use footprints derived from the three modelling approaches employed here are: i) Integrated Assessment Models – Energy Modelling Forum 33 (IAM EMF33); ii) Partial models assuming constant land cover (CLC), and, iii) Partial models using natural regrowth (NRG). The emissions factors used here correspond to scenarios where global production of biomass for energy purposes are 100 EJ/year, with lower emissions factors expected at lower levels of consumption and vice versa. Further details are available in Box 10.2 and Chapter 7.

The tailpipe emissions and fuel consumption reported in the literature generally do not use empirical emissions data. Rather, they tend to report fuel efficiency using driving cycles such as New European Driving Cycle or the US Environmental Protection Agency Federal Test Procedure. As a result, depending on the driving cycle used, operating emissions reported in literature are possibly underestimated by as much as 15–38%, in comparison to real driving emissions (Fontaras et al. 2017; Tsiakmakis et al. 2017; Triantafyllopoulos et al. 2019). The extent of these underestimations, however, varies between powertrain types, engine sizes, driving behaviour and environment.

Current average lifecycle impacts of mid-size ICEVs span from approximately 65 gCO2-eq pkm –1 to 210 gCO2-eq pkm –1, with both values stemming from ICEVs running on biofuels. Between this range of values, the current reference technologies are found, with diesel-powered ICEVs having total median lifecycle impacts of 130 gCO2-eq pkm –1 and gasoline-fuelled vehicle 160 gCO2-eq pkm –1. Fuel consumption dominates the lifecycle emissions of ICEVs, with approximately 75% of these emissions arising from the tailpipe and fuel chain.

HEVs and plug-in HEVs (PHEVs) vary in terms of degree of powertrain electrification. HEVs mainly rely on regenerative braking for charging the battery. PHEVs combine regenerative braking with external power sources for charging the battery. Operating emissions intensity is highly dependent on the degree to which electrified driving is performed, which in turn is user- and route-dependent. For PHEVs, emissions intensity is also dependent on the source of the electricity for charging. HEV and PHEV production impacts are comparable to the emissions generated for producing ICEVs as the batteries are generally small compared to those of BEVs. Current HEVs may reduce emissions compared to ICEVs by up to 30%, depending on the fuel, yielding median lifecycle intensities varying between 60 gCO2-eq pkm –1 (biofuels, EMF33) and 165–170 gCO2-eq pkm –1 (biofuels, partial models NRG). Within this wide range, all the combinations of electric and fossil-fuelled driving can be found, as well as the lifecycle intensity for driving 100% on fossil fuel. Because HEVs rely on combustion as the main energy conversion process, they offer limited mitigation opportunities. However, HEVs represent a suitable temporary solution, yielding a moderate mitigation potential, in areas where the electricity mix is currently so carbon intensive that the use of PHEVs and BEVs is not an effective mitigation solution (Wolfram and Wiedmann 2017; Wu et al. 2019).

In contrast to HEVs, PHEVs may provide greater opportunities for use-phase emissions reductions for LDVs. These increased potential benefits are due to the ability to charge the battery with low-carbon electricity and the longer full-electric range in comparison to HEVs (Laberteaux et al. 2019). Consumer behaviour (e.g., utility factor (UF) and charging patterns), manufacturer settings, and access to renewable electricity for charging strongly influence the total operational impacts (Wu et al. 2019). The UF is a weighting of the percentage of distance covered using the electric charge (charge depleting (CD) stage) versus the distance covered using the internal combustion engine (charge sustaining (CS) stage) (Paffumi et al. 2018). When the PHEV operates in CS mode, the internal combustion engine is used for propulsion and to maintain the state of charge of the battery within a certain range, together with regenerative braking (Plötz et al. 2018; Raghavan and Tal 2020). When running in CS mode, PHEVs have a reduced mitigation potential and have impacts comparable to those of HEVs. On the other hand, when the PHEV operates in CD mode, the battery alone provides the required propulsion energy (Plötz et al. 2018; Raghavan and Tal 2020). Thus, in CD mode, PHEVs hold potential for higher mitigation potential, due to the possibility of charging the battery with low-carbon electricity sources. Consequently, the UF greatly influences the lifecycle emissions of PHEVs. The current peer-reviewed literature presents a wide range of UFs mainly due to varying testing protocols applied for estimating the fuel efficiency and user behaviour (Pavlovic et al. 2017; Paffumi et al. 2018; Plötz et al. 2018; Plötz et al. 2020; Raghavan and Tal 2020; Hao et al. 2021). These factors make it difficult to harmonise and compare impacts across PHEV studies. Due to the low number of appropriate PHEV studies relative to the other LDV technologies and the complications in harmonising available PHEV results, this technology is omitted from Figure 10.4. However, due to the dual operating nature of PHEV vehicles, one can expect that the lifecycle GHG emissions intensities for these vehicles will lie between those of their ICEV and BEV counterparts of similar size and performance.

Currently, BEVs have higher manufacturing emissions than equivalently-sized ICEVs, with median emissions of 14 tCO2-eq per vehicle against approximately 10 tCO2-eq per vehicle of their mid-sized fossil-fuelled counterparts. These higher production emissions of BEVs are largely attributed to the battery pack manufacturing and to the additional power electronics required. As manufacturing technology and capacity utilisation improve and globalise to regions with low-carbon electricity, battery manufacturing emissions will likely decrease. Due to the higher energy efficiency of the electric powertrain, BEVs may compensate for these higher production emissions in the driving phase. However, the mitigation ability of this technology relative to ICEVs is highly dependent on the electricity mix used to charge the vehicle. As a consequence of the variety of energy sources available today, current BEVs have a wide range of potential average lifecycle impacts, ranging between 60 and 180 gCO2-eq pkm –1 with electricity generated from wind and coal, respectively. The ability to achieve large carbon reductions via vehicle electrification is thus highly dependent on the generation of low-carbon electricity, with the greatest mitigation effects achieved when charging the battery with low-carbon electricity. The literature suggests that current BEVs, if manufactured on low-carbon electricity as well as operated on low-carbon electricity would have footprints as low 22 gCO2-eq pkm –1 for a compact-sized car (Ellingsen et al. 2014; Ellingsen et al. 2016). This value suggests a reduction potential of around 85% compared to similarly-sized fossil fuel vehicles (median values). Furthermore, BEVs have a co-benefit of reducing local air pollutants that are responsible for human health complications, particularly in densely-populated areas (Hawkins et al. 2013; Ke et al. 2017).

As with BEVs, current HFCVs have higher production emissions than similarly-sized ICEVs and BEVs, generating on average approximately 15 tCO2-eq per vehicle. As with BEVs, the lifecycle impacts of FCVs are highly dependent on the fuel chain. To date, the most common method of hydrogen production is steam methane reforming of natural gas (Khojasteh Salkuyeh et al. 2017), which is relatively carbon intensive, resulting in lifecycle emissions of approximately 88 gCO2-eq pkm –1. Current literature covering lifecycle impacts of FCVs shows that vehicles fuelled with hydrogen produced from steam methane reforming of natural gas offer little or no mitigation potential over ICEVs. Other available hydrogen fuel chains vary widely in carbon intensity, depending on the synthesis method and the energy source used (electrolysis or steam methane reforming; fossil fuels or renewables). The least carbon-intensive hydrogen pathways rely on electrolysis powered by low-carbon electricity. Compared to ICEVs and BEVs, FCVs for LDVs are at a lower technology readiness level, as discussed in section 10.3.

Two-wheelers, consisting mainly of lower-powered mopeds and higher-powered motorcycles, are popular for personal transport in densely populated cities, especially in developing countries. LCA studies for this class of vehicle are relatively uncommon compared to four-wheeled LDVs. In the available results, however, two-wheelers exhibit similar trends for the different powertrain technologies as the LDVs, with electric powertrains having higher production emissions, but usually lower operating emissions. The lifecycle emissions intensity for two-wheelers is also generally lower than four-wheeled LDVs on a vehicle-kilometre basis. However, two-wheelers generally cannot carry as many passengers as four-wheeled LDVs. Thus, on a passenger-kilometre basis, a fully occupied passenger vehicle may still have lower emissions than a fully occupied two-wheeler. However, today, most passenger vehicles have relatively low occupancy and thus have a correspondingly high emissions intensity on a pkm basis. This points to the importance of utilisation of passenger vehicles at higher occupancies to reduce the lifecycle intensity of LDVs on a pkm basis. For example, the median emissions intensity of a gasoline passenger vehicle is 222 gCO2-eq vkm –1, and 160 gCO2-eq vkm –1 for a gasoline two-wheeler (Cox and Mutel 2018). At a maximum occupancy factor of four and two passengers, respectively, the transport emissions intensity for these vehicles is 55 and 80 gCO2-eq pkm –1. Under the same occupancy rates assumption, BEV two-wheelers recharged on the average European electricity mix, achieve lower lifecycle GHG intensities than BEV four-wheeled LDVs. On the other hand, FCV two-wheelers with hydrogen produced via steam methane reforming present higher GHG intensity than their four-wheeled counterparts, when compared on a pkm basis at high occupancy rates.

ICEV, HEV, and PHEV technologies, which are powered using combustion engines, have limited potential for deep reduction of GHG emissions. Biofuels offer good mitigation potential if low land-use change emissions are incurred (e.g., the IAM EMF33 and partial models, CLC biofuels pathways shown in Figure 10.4). The literature shows large variability, depending on the method of calculating associated land-use changes. Resolving these apparent methodological differences is important to consolidating the role biofuels may play in mitigation, as well as the issues raised in Chapter 7 about the conflicts over land use. The mitigation potential of battery and fuel cell vehicles is strongly dependent on the carbon intensity of their production and the energy carriers used in operation. However, these technologies likely offer the highest potential for reducing emissions from LDVs. Prior work on the diffusion dynamics of transport technologies suggests that ‘the diffusion of infrastructure precedes the adoption of vehicles, which precedes the expansion of travel’ (Leibowicz 2018). These dynamics reinforce the argument for strong investments in both the energy infrastructure and the vehicle technologies.

To successfully transition towards LDVs utilising low-carbon fuels or energy sources, the technologies need to be accessible to as many people as possible, which requires competitive costs compared to conventional diesel and gasoline vehicles. The lifecycle costs (LCCs) of LDVs depend on the purchasing costs of the vehicles, their efficiency, the fuel costs, and the discount rate. Figure 10.5 shows the results of a parametric analysis of LCC for diesel LDVs, BEVs, and FCVs. The range of vehicle efficiencies captured in Figure 10.5 is the same as the range used for Figure 10.4, while the ranges for fuel costs and vehicle purchase prices come from the literature. The assumed discount rate for this parametric analysis is 3%. Appendix 10.2 includes the details about the method and underlying data used to create this figure.

Figure 10.5 shows the range of LCC, in USD per passenger-kilometre, for different powertrain technologies, and the influence of vehicle efficiency (low or high), vehicle purchase price, and fuel/electricity cost on the overall LCC. For consistency with Figure 10.4, an occupancy rate of 1.5 is assumed. Mid-sized ICEVs have a purchase price of USD20,000–40,000, and average fuel costs are in the range of USD1–1.5 per litre. With these conditions, the LCC of fossil-fuelled LDVs span between USD0.22–0.35 pkm –1 or between USD0.17–0.28 pkm –1, for low- and high-efficiency ICEVs respectively (Figure 10.5).

BEVs have higher purchase prices than ICEVs, though a sharp decline has been observed since AR5. Due to the rapid development of the lithium-ion battery technology over the years (Schmidt et al. 2017) and the introduction of subsidies in several countries, BEVs are quickly reaching cost parity with ICEVs. Mid-sized BEVs’ average purchase prices are in the range of USD30,000–50,000 but the levelised cost of electricity shows a larger spread (USD65–200/MWh) depending on the geographical location and the technology (Chapter 6). Therefore, assuming purchase price parity between ICEVs and BEVs, BEVs show lower LCC (Figure 10.5) due to higher efficiency and the lower cost of electricity compared to fossil fuels on a per-gigajoule (GJ) basis (secondary y-axis on Figure 10.5).

FCVs represent the most expensive solution for LDV, mainly due to the currently higher purchase price of the vehicle itself. However, given the lower technology readiness level of FCVs and the current efforts in the research and development of this technology, FCVs could become a viable technology for LDVs in the coming years. The issues regarding the extra energy involved in creating the hydrogen and its delivery to refuelling sites remain, however. The levelised cost of hydrogen on a per GJ basis is lower than conventional fossil fuels but higher than electricity. In addition, within the levelised cost of hydrogen, there are significant cost differences between the hydrogen-producing technologies. Conventional technologies such as coal gasification and steam methane reforming of natural gas, both with and without carbon capture and storage, represent the cheapest options (Bekel and Pauliuk 2019; Parkinson et al. 2019; Khzouz et al. 2020; Al-Qahtani et al. 2021). Hydrogen produced via electrolysis is currently the most expensive technology, but with significant potential cost reductions due to the current technology readiness level.

Buses provide urban and peri-urban transport services to millions of people around the world and a growing number of transport agencies are exploring alternative-fuelled buses. Alternative technologies to conventional diesel-powered buses include buses powered with CNG, LNG, synthetic fuels, and biofuels (e.g., biodiesel, renewable diesel, dimethyl ether); diesel hybrid-electric buses; battery electric buses; electric catenary buses; and hydrogen fuel cell buses. Rail is an alternative mode of transit that could support decarbonisation of land-based passenger mobility. Electric rail systems can provide urban services (light rail and metro systems), as well as longer-distance transport. Indeed, many cities of the world already have extensive metro systems, and regions like China, Japan and Europe have a robust high-speed intercity railway network. Intercity rail transport can be powered with electricity, however, fossil fuels are still prevalent for long-distance rail passenger transport in some regions. Battery electric long-distance trains may be a future option for these areas.

Figure 10.6 shows the lifecycle GHG emissions from different powertrain and fuel technologies for buses and passenger rail. The data in each panel came from a number of relevant scientific studies (Cai et al. 2015; Tong et al. 2015a; Dimoula et al. 2016; de Bortoli et al. 2017; Valente et al. 2017; Meynerts et al. 2018; IEA 2019e; de Bortoli and Christoforou 2020; Hill et al. 2020; Liu et al. 2020a; Valente et al. 2021). The width of the bar represents the variabilityin available estimates, which is primarily driven by variability in reported vehicle efficiency, size, or drive cycle. While some bars overlap, the Figure may not fully capture correlations between results. For example, low efficiency associated with aggressive drive cycles may drive the upper end of the emission ranges for multiple technologies; thus, an overlap does not necessarily suggest uncertainty regarding which vehicle type would have lower emissions for a comparable trip. Additionally, reported lifecycle emissions do not include embodied GHG emissions associated with infrastructure construction and maintenance. These embodied emissions are potentially a larger fraction of lifecycle emissions for rail than for other transport modes (Chester and Horvath 2012; Chester et al. 2013). One study reported values ranging from 10–25 gCO2 per passenger-kilometre (International Union of Railways 2016), although embodied emissions from rail are known to vary widely across case studies (Olugbenga et al. 2019). These caveats are also applicable to the other figures in this section.

Figure 10.6 highlights that BEV and FCV buses and passenger rail powered with low-carbon electricity or low-carbon hydrogen, could offer reductions in GHG emissions compared to diesel-powered buses or diesel-powered passenger rail. However, and not surprisingly, these technologies would offer only little emissions reductions if power generation and hydrogen production rely on fossil fuels. While buses powered with CNG and LNG could offer some reductions compared to diesel-powered buses, these reductions are unlikely to be sufficient to contribute to deep decarbonisation of the transport sector and they may slow down conversion to low- or zero-carbon options already commercially available. Biodiesel and renewable diesel fuels (from sources with low upstream emissions and low risk of induced land-use change) could offer important near-term reductions for buses and passenger rail, as these fuels can often be used with existing vehicle infrastructure. They could also be used for long haul trucks and trains, shipping and aviation as discussed below and in later sections.

There has been growing interest in the production of synthetic fuels from CO2 produced by direct air capture (DAC) processes. Figure 10.6 includes the lifecycle GHG emissions from buses and passenger rail powered with synthetic diesel produced through a DAC system paired with a Fischer-Tropsch (FT) process, based on Liu et al. (2020a). This process requires the use of hydrogen (as shown in Figure 10.2), so the emissions factors of the resulting fuel depend on the emissions intensity of hydrogen production. An electricity emissions factor less than 140 gCO2-eq kWh –1 would be required for this pathway to achieve lower emissions than petroleum diesel (Liu et al. 2020a); for example, this would be equivalent to a 75% wind and 25% natural gas electricity mix (Appendix 10.1). If the process relied on steam methane reforming for hydrogen production or fossil-based power generation, synthetic diesel from the DAC-FT process would not provide GHG emissions reductions compared to conventional diesel. DAC-FT from low-carbon energy sources appears to be promising from an emissions standpoint and could warrant the R&D and demonstration attention outlined in the rest of the chapter, but it cannot be contemplated as a decarbonisation strategy without the availability of low-carbon hydrogen.

At high occupancy, both bus and rail transport offer substantial GHG reduction potential per pkm, even compared with the lowest-emitting private vehicle options. Even at 20% occupancy, bus and rail may still offer emission reductions compared to passenger cars, especially notable when comparing BEVs with low-carbon electricity (the lowest-emission option for all technologies) across the three modes. Only when comparing a fossil fuel-powered bus at low occupancy with a low-carbon powered car at high occupancy is this conclusion reversed. Use of public transit systems, especially those that rely on buses and passenger rail fuelled with the low-carbon fuels previously described, would thus support efforts to decarbonise the transport sector. Use of these public transit systems will depend on urban design and consumer preferences (Section 10.2, Chapters 5 and 8), which in turn depend on time, costs, and behavioural choices.

Figure 10.7 shows the results of a parametric analysis of the LCCs of transit technologies with the highest potential for GHG emissions reductions. As with Figure 10.5, the vehicle efficiency ranges are the same as those from the LCA estimates (80% occupancy). Vehicle, fuel, and maintenance costs represent ranges in the literature (Eudy and Post 2018b; IEA 2019e; Argonne National Laboratory 2020; BNEF 2020; Eudy and Post 2020; Hydrogen Council 2020; IEA 2020b; IEA 2020c; IRENA 2020; Johnson et al. 2020; Burnham et al. 2021; IEA 2021c; IEA 2021d; US Energy Information Administration 2021), and the discount rate is 3% where applicable. Appendix 10.2 provides the details behind these estimates. The panels for the ICEV can represent buses and passenger trains powered with any form of diesel, whether derived from petroleum, synthetic hydrocarbons, or biofuels. For reference, global average automotive diesel prices from 2015–2020 fluctuated around USD1 per litre, and the 2019 world average industrial electricity price was approximately USD100 per MWh (IEA 2021d). Retail hydrogen prices in excess of USD13 per kilogram have been observed(Eudy and Post 2018a; Argonne National Laboratory 2020; Burnham et al. 2021) though current production cost estimates for hydrogen produced from electrolysis are far lower (IRENA 2020) (and as reported in Chapter 6), at around USD5–7 per kg with future forecasts as low as USD1 per kg (BNEF 2020; Hydrogen Council 2020; IRENA 2020) (and as reported in Chapter 6).

Under most parameter combinations, rail is the most cost-effective option, followed by buses, both of which are an order of magnitude cheaper than passenger vehicles. Note that costs per pkm are strongly influenced by occupancy assumptions; at low occupancy (e.g., <20% for buses and <10% for rail), the cost of transit approaches the LCC for passenger cars. For diesel rail and buses, cost ranges are driven by fuel costs, whereas vehicles are both important drivers for electric or hydrogen modes due to high costs (but also large projected improvements) associated with batteries and fuel cell stacks. Whereas the current state of ICEV technologies is best represented by cheap vehicles and low fuel costs for diesel (top left of each panel), these costs are likely to rise in future due to stronger emission/efficiency regulations and rising crude oil prices. On the contrary, the current status of alternative fuels is better represented by high capital costs and mid-to-high fuel costs (right side of each panel; mid-to-bottom rows), but technology costs are anticipated to fall with increasing experience, research, and development. Thus, while electric rail is already competitive with diesel rail, and electric buses are competitive with diesel buses in the low efficiency case, improvements are still required in battery costs to compete against modern diesel buses on high efficiency routes, at current diesel costs. Similarly, improvements to both vehicle cost and fuel costs are required for hydrogen vehicles to become cost effective compared to their diesel or electric counterparts. At either the upper end of the diesel cost range (bottom row of ICEV panels), or within the 2030–2050 projections for battery costs, fuel cell costs and hydrogen costs (top left of BEV and FCV panels), both battery- and hydrogen-powered vehicles become financially attractive.

As is the case with passenger transport, thereis growing interest in alternative fuels that could reduce GHG emissions from freight transport. Natural gas-based fuels (e.g., CNG, LNG) are an example, however these may not lead to drastic reductions in GHG emissions compared to diesel. Natural gas-powered vehicles have been discussed as a means to mitigate air quality impacts (Khan et al. 2015; Cai et al. 2017; Pan et al. 2020), but those impacts are not the focus of this review. Decarbonisation of medium- and heavy-duty trucks would likely require the use of low-carbon electricity in battery electric trucks, low-carbon hydrogen or ammonia in fuel-cell trucks, or bio-based fuels (from sources with low upstream emissions and low risk of induced land-use change) used in ICE trucks.

Freight rail is also a major mode for the inland movement of goods. Trains are more energy efficient (per tkm) than trucks, so expanded use of rail systems (particularly in developing countries where demand for goods could grow exponentially) could provide carbon abatement opportunities. While diesel-based locomotives are still a major mode of propulsion used in freight rail, interest in low-carbon propulsion technologies is growing. Electricity already powers freight rail in many European countries using overhead catenaries. Other low-carbon technologies for rail may include advanced storage technologies, biofuels, synthetic fuels, ammonia, or hydrogen.

Figure 10.8 presents a review of lifecycle GHG emissions from land-based freight technologies (heavy- and medium-duty trucks, and rail). Each panel within the figure represents data in GHG emissions per tonne-kilometre of freight transported by different technology and/or fuel types, as indicated by the labels to the left. The data in each panel came from a number of relevant scientific studies (Tong et al. 2015a; Frattini et al. 2016; Nahlik et al. 2016; Zhao et al. 2016; CE Delft 2017; Isaac and Fulton 2017; Song et al. 2017; Valente et al. 2017; Cooper and Balcombe 2019; Lajevardi et al. 2019; Hill et al. 2020; Liu et al. 2020a; Merchan et al. 2020; Prussi et al. 2020; Gray et al. 2021; Valente et al. 2021). Similar to the results for buses, technologies that offer substantial emissions reductions for freight include: ICEV trucks powered with the low-carbon variants for biofuels, ammonia or synthetic diesel; BEVs charged with low-carbon electricity; and FCVs powered with renewable-based electrolytic hydrogen, or ammonia. Since ammonia and Fischer-Tropsch diesel are produced from hydrogen, their emissions are higher than the source hydrogen, but their logistical advantages over hydrogen are also a consideration (Section 10.3).

Trucks exhibit economies of scale in fuel consumption, with heavy-duty trucks generally showing lower emissions per tkm than medium-duty trucks. Comparing the lifecycle GHG emissions from trucks and rail, it is clear that rail using internal combustion engines is more carbon efficient than using internal combustion trucks. Note that the rail emissions are reported for an average representative payload, while the trucks are presented at 50% and 100% payload, based on available data. The comparison between trucks and rail powered with electricity or hydrogen is less clear – especially considering that these values omit embodied GHG from infrastructure construction. One study reported embodied rail infrastructure emissions of 15 gCO2 per tonne-kilometre for rail (International Union of Railways 2016), although such embodied emissions from rail are known to vary widely across case studies (Olugbenga et al. 2019). Regardless, trucks and rail with low-carbon electricity or low-carbon hydrogen have substantially lower emissions than incumbent technologies.

For trucks, Figure 10.8 includes two x-axes representing two different assumptions about their payload, which substantially influence emissions per tonne-kilometre. These results highlight the importance of truckload planning as an emissions reduction mechanism, for example, as also shown in Kaack et al. (2018). Several studies also point to improvements in vehicle efficiency as an important mechanism to reduce emissions from freight transport (Taptich et al. 2016; Kaack et al. 2018). However, projections for diesel vehicles using such efficiencies beyond 2030 are promising, but still far higher emitting than vehicles powered with low-carbon sources.

Figure 10.9 shows the results of a parametric analysis of the LCC of trucks and freight rail technologies with the highest potential for deep GHG reductions. As with Figure 10.8, the vehicle efficiency ranges are the same as those from the LCA estimates (80% payload for trucks; effective payload as reported by original studies for rail). Vehicle, fuel and maintenance costs represent ranges in the literature (Moultak et al. 2017; Eudy and Post 2018b; IEA 2019e; Argonne National Laboratory 2020; BNEF 2020; IRENA 2020; Burnham et al. 2021; IEA 2021c), and the discount rate is 3% where applicable (Appendix 10.2). The panels for the ICEV can represent trucks and freight trains powered with any form of diesel, whether derived from petroleum, synthetic hydrocarbons, or biofuels. See discussion preceding Figure 10.7 for additional details about current global fuel costs. Under most parameter combinations, rail is the more cost-effective option, but the high efficiency case for trucks (representing fuel-efficient vehicles, favourable drive cycles and high payload) can be more cost-effective than the low efficiency case for rail (representing systems with higher fuel consumption and lower payload). For BEV trucks, cost ranges are driven by vehicle purchase price due to the large batteries required and the associated wide range between their current high costs and anticipated future cost reductions. For all other truck and rail technologies, fuel cost ranges play a larger role. Similar to transit technologies, the current state of freight ICEV technologies is best represented by cheap vehicles and low fuel costs for diesel (top left of each panel), and the current status of alternative fuels is better represented by high capital costs and mid-to-high fuel costs (right side of each panel; mid-to-bottom rows), with expected future increases in ICEV LCC and decreases in alternative fuel vehicle LCC. Electric and hydrogen freight rail are potentially already competitive with diesel rail (especially electric catenary (IEA 2019e)), but low data availability (especially for hydrogen efficiency ranges) and wide ranges for reported diesel rail efficiency (likely encompassing low capacity utilisation) makes this comparison challenging. Alternative fuel trucks are currently more expensive than diesel trucks, but future increases in diesel costs or a respective decrease in hydrogen costs or in BEV capital costs (especially the battery) would enable either alternative fuel technology to become financially attractive. These results are largely consistent with raw results reported in existing literature, which suggest ambiguity over whether BEV trucks are already competitive, but more consistency that hydrogen is not yet competitive, but could be in future (Zhao et al. 2016; Moultak et al. 2017; Sen et al. 2017; White and Sintov 2017; Zhou et al. 2017; Mareev et al. 2018; Yang et al. 2018a; El Hannach et al. 2019; Lajevardi et al. 2019; Tanco et al. 2019; Burke and Sinha 2020; Jones et al. 2020). There is limited data available on the LCC for freight rail, but at least one study IEA (2019g) suggests that electric catenary rail is likely to have similar costs to diesel rail, while battery electric trains remain more expensive and hydrogen rail could become cheaper under forward-looking cost reduction scenarios.

Taken together, the results in this section suggest a range of cost-effective opportunities to reduce GHG emissions from land-based transport. Mode shift from cars to passenger transit (bus or rail) can reduce GHG emissions while also reducing LCCs, resulting in a negative abatement cost. Likewise, increasing the utilisation of vehicles (i.e., % occupancy for passenger vehicles or % payload for freight vehicles) simultaneously decreases emissions and costs per pkm or per tkm, respectively. Within a given mode, alternative fuel sources also show strong potential to reduce emissions at minimal added costs. For LDVs, BEVs can offer emissions reductions with LCCs that are already approaching that for conventional ICEVs. For transit and freight, near-term abatement costs for the low-carbon BEV and FCV options relative to their diesel counterparts range from near USD0/tonne CO2-eq (e.g., BEV buses and BEV passenger rail) into the hundreds or even low thousands of dollars per tonne CO2-eq (e.g., for heavy-duty BEV and FCV trucks at current vehicle and fuel costs). With projected future declines in storage, fuel cell, and low-carbon hydrogen fuel costs, however, both BEV and FCV technologies can likewise offer GHG reductions at negative abatement costs across all land-transport modes in 2030 and beyond. Further information about costs and potentials is available in Chapter 12.

Aviation is widely recognised as a ‘hard-to-decarbonise’ sector (Gota et al. 2019) having a strong dependency on liquid fossil fuels and an infrastructure that has long ‘lock-in’ timescales, resulting in slow fleet turnover times. The principal GHG emitted is CO2 from the combustion of fossil fuel aviation kerosene (‘Jet-A’), although its non-CO­2 emissions can also affect climate (Section 10.5.2). International emissions of CO2 are about 65% of the total emissions from aviation (Fleming and de Lépinay 2019), which totalled approximately 1 Gt of CO2 in 2018. Emissions from this segment of the transport sector have been steadily increasing at rates of around 2.5% per year over the last two decades (Figure 10.10), although for the period 2010 to 2018 the rate increased to roughly 4% per year. The latest available data (2018) indicate that aviation is responsible for approximately 2.4% of total anthropogenic emissions of CO2 (including land-use change) on an annual basis (using IEA data, IATA data and global emissions data of Le Quéré et al. (2018b)).

Aviation’s net warming effect results from its historical and current emissions of CO2, and non-CO2 emissions of water vapour, soot, sulphur dioxide (from sulphur in the fuel), and nitrogen oxides (NOx, = NO + NO2) (IPCC 1999; Lee et al. 2021; Szopa et al. 2021). Although the effective radiative forcing (ERF) of CO2 from historic aviation emissions is not currently the largest forcing term, it is difficult to address because of the sector’s current dependency on fossil-based hydrocarbon fuels and the longevity of CO2. A residual of emissions of CO2 today will still have a warming effect in many thousands of years (Archer et al. 2009; Canadell et al. 2021) whereas water vapour, soot, and NOx emissions will have long ceased to contribute to warming after some decades. As a result, CO2 mitigation of aviation to net zero levels, as required in 1.5°C scenarios, requires fundamental shifts in technology, fuel types, or changes of behaviour or demand.

The non-CO2 effects of aviation on climate fall into the category of short-lived climate forcers (SLCFs). Emissions of NOx currently result in net positive warming from the formation of short-term ozone (warming) and the destruction of ambient methane (cooling). If the conditions are suitable, emissions of soot and water vapour can trigger the formation of contrails (Kärcher 2018), which can spread to form extensive contrail-cirrus cloud coverage. Such cloud coverage is estimated to have a combined ERF that is about 57% of the current net ERF of global aviation (Lee et al. 2021), although a comparison of cirrus cloud observations under pre- and post-COVID-19 pandemic conditions suggest that this forcing could be smaller (Digby et al. 2021). Additional effects from aviation from aerosol-cloud interactions on high-level ice clouds through soot (Chen and Gettelman 2013; Zhou and Penner 2014; Penner et al. 2018), and lower-level warm clouds through sulphur (Righi et al. 2013; Kapadia et al. 2016) are highly uncertain, with no best estimates available (Lee et al. 2021). In total, the net ERF from aviation’s non-CO2 SLCFs is estimated to be approximately 66% of aviation’s current total forcing. It is important to note that the fraction of non-CO2 forcing to total forcing is not a fixed quantity and is dependent on the recent history of growth (or otherwise) of CO2 emissions (Klöwer et al. 2021). The non-CO2 effects from aviation are the subject of discussion for mitigation options (Arrowsmith et al. 2020). However, the issues are complex, potentially involving technological and operational trade-offs with CO2.

Technology options for engine and airframe. For every kilogram of jet fuel combusted, 3.16 kg CO2 is emitted. Engine and airframe manufacturers’ primary objective, after safety issues, is to reduce direct operating costs, which are highly dependent on fuel burn. Large investments have gone into engine technology and aircraft aerodynamics to improve fuel burn per kilometre (Cumpsty et al. 2019). There have been major step changes in engine technology over time, from early turbojet engines to larger turbofan engines. However, the basic configuration of an aircraft has remained more or less the same for decades and will likely remain at least to 2037 (Cumpsty et al. 2019). Airframes performance has improved over the years with better wing design, but large incremental gains have become much harder as the technology has matured. For twin-aisle aircraft, generally used for long ranges, fuel-burn is a pressing concern and there have been several all-new aircraft designs with improvements in their lift-to-drag ratio (Cumpsty et al. 2019). The principal opportunities for fuel reduction come from improvements in aerodynamic efficiency, aircraft mass reduction, and propulsion system improvements. In the future, Cumpsty et al. (2019) suggest that the highest rate of fuel burn reduction achievable for new aircraft is likely to be no more than about 1.3% per year, which is well short of ICAO’s aspirational goal of 2% global annual average fuel efficiency improvement. Radically different aircraft shapes, like the blended wing body (where the wings are not distinct from the fuselage), are likely to use about 10% less fuel than future advanced aircraft of conventional form (Cumpsty et al. 2019). Such improvements would be ‘one-off’ gains, do not compensate for growth in emissions of CO2 expected to be in excess of 2% per annum, and would take a decade or more to penetrate the fleet completely. Thus, the literature does not support the idea that there are large improvements to be made in the energy efficiency of aviation that keep pace with the projected growth in air transport.

Operational improvements for navigation. From a global perspective, aircraft navigation is relatively efficient, with many long-haul routes travelling close to great circle trajectories, and avoiding headwinds that increase fuel consumption. The ICAO estimates that flight inefficiencies on a global basis are currently of the order 2% to 6% (ICAO 2019), while Fleming and de Lépinay (2019) project operational improvements (air traffic management) of up to 13% on a regional basis by 2050. ‘Intermediate stop operations’ have been suggested, whereby longer-distance travel is broken into flight legs, obviating the need to carry fuel for the whole mission. Linke et al. (2017) modelled this operational behaviour on a global basis and calculated a fuel saving of 4.8% over a base case in which normal fuel loads were carried. However, this approach increases the number of landing/take-off cycles at airports. ‘Formation flying’, which has the potential to reduce fuel burn on feasible routes, has also been proposed (Xu et al. 2014; Marks et al. 2021).

Alternative biofuels, synthetic fuels, and liquid hydrogen. As noted above, the scope for reducing CO2 emissions from aviation through improved airplane technology or operations is limited and unable to keep up with the projected growth, let al.ne reduce beyond the present emission rate at projected levels of demand (assuming post-pandemic recovery of traffic). Thus, the literature outlined here suggests that the only way for demand for aviation to continue to grow without increasing CO2 emissions is to employ alternative lower-carbon bio- or synthetic aviation fuels (Klöwer et al. 2021). For shorter ranges, flights of light planes carrying up to 50 passengers may be able to use electric power (Sahoo et al. 2020) but these planes are a small proportion of the global aviation fleet (Epstein and O’Flarity 2019; Langford and Hall 2020) and account for less than 12% of current aviation CO2 emissions. Alternative lower-carbon footprint fuels have been certified for use over recent years, principally from bio-feedstocks, but are not yet widely available at economic prices (Kandaramath Hari et al. 2015; Capaz et al. 2021a). In addition, alternative fuels from bio-feedstocks have variable carbon footprints because of different lifecycle emissions associated with various production methods and associated land-use change (de Jong et al. 2017; Staples et al. 2018; Capaz et al. 2021b; Zhao et al. 2021).

The development of ‘sustainable aviation fuels’ (referred to as ‘SAFs’) that can reduce aviation’s carbon footprint is a growing area of interest and research. Alternative aviation fuels to replace fossil-based kerosene have to be certified to an equivalent standard as Jet-A for a variety of parameters associated with safety issues. Currently, the organisation responsible for aviation fuel standards, ASTM International, has certified seven different types of sustainable aviation fuels with maximum blends ranging from 10% to 50% (Chiaramonti 2019). Effectively, these blend requirements limit the amount of non-hydrocarbon fuel (e.g., methanol) that can be added at present. While there currently is a minimum level of aromatic hydrocarbon contained in jet fuel to prevent ‘O-ring’ shrinkage in the fuel seals (Khandelwal et al. 2018), this minimum level can likely be lower in the medium to long term, with the added benefits of reduced soot formation and reduced contrail cirrus formation (Bier et al. 2017; Bier and Burkhardt 2019).

Bio-based fuels can be produced using a variety of feedstocks including cultivated feedstock crops, crop residues, municipal solid waste, waste fats, oils and greases, wood products and forestry residues (Staples et al. 2018). Each of these different sources can have different associated lifecycle emissions, such that they are not net zero CO2 emissions but have associated emissions of CO2 or other GHGs from their production and distribution (Section 10.3, Box 10.2). In addition, associated land-use change emissions of CO2 represent a constraint in climate change mitigation potential with biofuel (Staples et al. 2017) and have inherent large uncertainties (Plevin et al. 2010). Other sustainability issues include food vs fuel arguments, water resource use, and impacts on biodiversity. Cost-effective production, feedstock availability, and certification costs are also relevant (Kandaramath Hari et al. 2015). Nonetheless, bio-based SAFs have been estimated to achieve lifecycle emissions reductions ranging between approximately 2% and 70% under a wide range of scenarios (Staples et al. 2018). For a set of European aviation demand scenarios, Kousoulidou and Lonza (2016) estimated that the fuel demand in 2030 would be about 100 million tonnes of oil equivalent and biokerosene (HEFA/HVO) penetration would provide around 2% of the total fuel demand at that date. Several issues limit the expansion of biokerosene for aviation, the primary one being the current cost of fossil fuel compared to the costs of SAF production (Capaz et al. 2021a). Other hybrid pathways, for example the hydrogenation of biofuels (the hydrogen assumed to be generated with low-carbon energy), could increase the output and improve the economic feasibility of bio-based SAF (Hannula 2016; Albrecht et al. 2017).

Costs remain a major barrier for bio-SAF, which cost around three times the price of kerosene (Kandaramath Hari et al. 2015). Clearly, for SAFs to be economically competitive, large adjustments in prices of fossil fuels or the introduction of policies is required. Staples et al. (2018) estimated that in order to introduce bio-SAFs that reduce lifecycle GHG emissions by at least 50% by 2050, prices and policies were necessary for incentivisation. They estimate the need for 268 new biorefineries per year and capital investments of approximately 22 to 88 billion USD2015 per year between 2020 and 2050. Wise et al. (2017) suggest that carbon prices would help leverage production and availability.

Various pathways have been discussed for the production of non-bio SAFs such as power-to-liquid pathways (Schmidt et al. 2018), sometimes termed ‘electro-fuels’ (Goldmann et al. 2018), or more generalised ‘Power-to-X’ pathways (Kober and Bauer 2019). This process would involve the use of low-carbon electricity, CO2, and water to synthesise jet fuel through the Fischer-Tropsch process or methanol synthesis. Hydrogen would be produced via an electrochemical process, powered by low-carbon energy and combined with CO2 captured directly from the atmosphere or through BECCS. The energy requirement from photovoltaics has been estimated to be of the order 14 to 20 EJ to phase out aviation fossil fuel by 2050 (Gössling et al. 2021a). These synthetic fuels have potential for large lifecycle emissions reductions (Schmidt et al. 2016). In comparison to bio-SAF production, the implementation of the processes is in its infancy. However, assuming availability of low-carbon energy electricity, these fuels have much smaller land and water requirements than bio-SAF. Low carbon-energy supply, scalable technology, and therefore costs, represent barriers. Scheelhaase et al. (2019) review current estimates of costs, which are estimated to be approximately four to six times the price of fossil kerosene.

Liquid hydrogen (LH2) as a fuel has been discussed for aeronautical applications since the 1950s (Brewer 1991) and a few experimental aircraft have flown using such a fuel. Experimental, small aircraft have also flown using hydrogen fuel cells. Although the fuel has an energy density per unit mass about three times greater than kerosene, it has a much lower energy density per unit volume (approximately factor 4 (McKinsey 2020)). The increased volume requirement makes the fuel less attractive for aviation since it would require the wings to be thickened or fuel to take up space in the fuselage. Bicer and Dincer (2017) found that LH2-powered aircraft compared favourably to conventional kerosene-powered aircraft on a lifecycle basis, providing that the LH2 was generated from low-carbon energy sources (0.014 kgCO2 per tkm compared with 1.03 kgCO2 per tkm for an unspecified passenger aircraft). However, Ramos Pereira et al. (2014) also made a lifecycle comparison and found much smaller benefits of LH2-powered aircraft (manufactured from low-carbon energy) compared with conventional fossil kerosene. The two studies expose the sensitivities of boundaries and assumptions in the analyses. Shreyas Harsha (2014) and Rondinelli et al. (2017) conclude that there are many infrastructural barriers but that the environmental benefits of low-carbon-based LH­2 could be considerable. Khandelwal et al. (2013) take a more optimistic view of the prospect of LH2-powered aircraft but envisage them within a hydrogen-oriented energy economy. A recently commissioned study by the European Union (EU)’s Clean Sky undertaking, (McKinsey 2020) addresses many of the aspects of the opportunities and obstacles in developing LH2-powered aircraft. The report provides an optimistic view of the feasibility of developing such aircraft for short to medium haul but makes clear that new aircraft designs (such as blended-wing body aircraft) would be needed for longer distances.

The non-CO2 impacts of LH2-powered aircrafts remain poorly understood. The emission index of water vapour would be much larger (estimated to be 2.6 times greater by Ström and Gierens (2002)) than for conventional fuels), and the occurrence of contrails may increase but have lower ERF because of the lower optical depth (Marquart et al. 2005). Moreover, contrails primarily form on soot particles from kerosene-powered aircraft, which would be absent from LH2 exhaust (Kärcher 2018). The overall effect is currently unknown as there are no measurements. Potentially, NOx emissions could be lower with combustor redesign (Khandelwal et al. 2013).

In conclusion, there are favourable arguments for LH2-powered aircraft, both on an efficiency basis (Verstraete 2013) and an overall reduction in GHG emissions, even on an lifecycle basis. However, LH2 requires redesign of the aircraft, particularly for long-haul operations. Similarly, there would be a need for expanded infrastructure for fuel manufacture, storage, and distribution at airports, which is likely to be more easily overcome if there is a more general move towards a hydrogen-based energy economy.

Technological and operational trade-offs between CO2and non-CO2effects. Since aviation has additional non-CO2 warming effects, there has been some discussion as to whether these can be addressed by either technological or operational means. For example, improved fuel efficiency has resulted from high overall pressure ratio engines with large bypass ratios. This improvement has increased pressure and temperature at the combustor inlet, with a resultant tendency to increase thermal NOx formation in the combustor. Combustor technology aims to reduce this increase, but it represents a potential technology trade-off whereby NOx control may be at the expense of extra fuel efficiency. Estimating the benefits or disbenefits of CO2 (proportional to fuel burned) vs NOx in terms of climate is complex (Freeman et al. 2018).

Any global warming potential/temperature change potential type emissions equivalency calculation always involves the user selection of a time horizon over which the calculation is made, which is a subjective choice (Fuglestvedt et al. 2010). In general, the longer the time horizon, the more important CO2 becomes in comparison with a short-lived climate forcing agent. So, for example, a net (overall) aviation GWP for a 20-year time horizon is 4.0 times that of CO2 alone, but only 1.7 over a 100-year time horizon. Correspondingly, a GTP for a 20-year time horizon is 1.3, but it is 1.1 for 100 years (Lee et al. 2021).

A widely discussed opportunity for mitigation of non-CO2 emissions from aviation is the avoidance of persistent contrails that can form contrail cirrus. Contrails only form in ice-supersaturated air below a critical temperature threshold (Kärcher 2018). It is therefore feasible to alter flight trajectories to avoid such areas conducive to contrail formation, since ice-supersaturated areas tend to be tens to hundreds of kilometres in the horizontal and only a few 100 metres in the vertical extent (Gierens et al. 1997). Theoretical approaches show that avoidance is possible on a flight-by-flight basis (Matthes et al. 2017; Teoh et al. 2020). Case studies have shown that flight planning according to trajectories with minimal climate impact can substantially (up to 50%) reduce the aircraft’s net climate impacts despite small additional CO2 emissions (Niklaß et al. 2019). However, any estimate of the net benefit or disbenefit depends firstly on the assumed magnitude of the contrail cirrus ERF effect (itself rather uncertain, assessed with a low confidence level) and upon the choice of metric and time horizon applied. While this is a potentially feasible mitigation option, notwithstanding the CO2 per contrail trade-off question, meteorological models cannot currently predict the formation of persistent contrails with sufficient accuracy in time and space (Gierens et al. 2020); this mitigation option is speculated to take of the order of up to a decade to mature (Arrowsmith et al. 2020).

Market-based offsetting measures. The EU introduced aviation into its CO2 emissions trading scheme (ETS) in 2012. Currently, the EU-ETS for aviation includes all flights within the EU as well as to and from Eastern European and West-Central Asian states. Globally, ICAO agreed in 2016 to commence, in 2020, the ‘Carbon Offsetting and Reduction Scheme for International Aviation’ (CORSIA). The pandemic subsequently resulted in the baseline being changed to 2019.

CORSIA has a phased implementation, with an initial pilot phase (2021–2023) and a first phase (2024–2026) in which states will participate voluntarily. The second phase will then start in 2026–2035, and all states will participate unless exempted. States may be exempted if they have lower aviation activity levels or based on their UN development status. As of September 2021, 109 ICAO Member States will voluntarily be participating in CORSIA starting in 2022. In terms of routes, only those where both States connecting the route are participating are included. There will be a special review of CORSIA by the end of 2032 to determine the termination of the scheme, its extension, or any other changes to the scheme beyond 2035.

By its nature, CORSIA does not lead to a reduction in in-sector emissions from aviation since the programme deals mostly in approved offsets. At its best, CORSIA is a transition arrangement to allow aviation to reduce its impact in a more meaningful way later. From 2021 onwards, operators can reduce their CORSIA offsetting requirements by claiming emissions reductions from ‘CORSIA Eligible Fuels’ that have demonstrably reduced lifecycle emissions. These fuels are currently available at greater costs than the offsets (Capaz et al. 2021a). As a result, most currently approved CORSIA offsets are avoided emissions, which raises the issue of additionality (Warnecke et al. 2019). The nature of avoided emissions is to prevent an emission that was otherwise considered to be going to occur, for example, prevented deforestation. Avoided emissions are ‘reductions’ (over a counterfactual) and purchased from other sectors that withhold from an intended emission (Becken and Mackey 2017), such that if additionality were established, a maximum of 50% of the intended emissions are avoided. Some researchers suggest that avoided deforestation offsets are not a meaningful reduction, since deforestation continues to be a net source of CO2 emissions (Mackey et al. 2013; Friedlingstein et al. 2020).

Modal shift to high-speed rail. Due to the limitations of the current suite of aviation mitigation strategies, the potential for high-speed rail (HSR) is of increasing interest (Givoni and Banister 2006; Chen 2017; Bi et al. 2019). The IEA’s Net Zero by 2050 roadmap suggests significant behavioural change, with more regional flights shifting to HSR in the Net Zero Emissions by 2050 scenario pathway (IEA 2021e). For HSR services to be highly competitive with air travel, the optimal distance between the departure and arrival points has been found to be in the approximate range of 400 to 800 km (Bows et al. 2008; Rothengatter 2010), although in the case of China’s HSR operations, this range can be extended out to 1000 km, with corresponding air services having experienced significant demand reduction upon HSR service commencement (Lawrence et al. 2019). In some instances, negative effects on air traffic, air fare, and flight frequency have occurred at medium-haul distances such as HSR services in China on the Wuhan–Guangzhou route (1069 km) and the Beijing–Shanghai route (1318 km) (Fu et al. 2015; Zhang and Zhang 2016; Chen 2017; Li et al. 2019; Ma et al. 2019). This competition at medium-haul distances is contrary to that which has been experienced in European and other markets and may be attributable to China having developed a comprehensive network with hub stations, higher average speeds, and an integrated domestic market with strong patronage (Zhang et al. 2019a).

The LCA literature suggests that the GHG emissions associated with HSR vary depending on spatial, temporal, and operational specifics (Åkerman 2011; Baron et al. 2011; Chester and Horvath 2012; Yue et al. 2015; Hoyos et al. 2016; Jones et al. 2017; Robertson 2016; Robertson 2018; Lin et al. 2019). These studies found a wide range of approximately 10 to 110 gCO2 pkm –1 for HSR. This range is principally attributable to the sensitivity of operational parameters such as the HSR passenger seating capacity, load factor, composition of renewable and non-renewable energy sources in electricity production, rolling stock energy efficiency and patronage (i.e., ridership both actual and forecast), and line-haul infrastructure specifics (e.g., tunnelling and aerial structure requirements for a particular corridor) (Åkerman 2011; Chester and Horvath 2012; Yue et al. 2015; Newman et al. 2018; Robertson 2018). The prospect for HSR services providing freight carriage (especially online purchases) is also growing rapidly (Strale 2016; Bi et al. 2019; Liang and Tan 2019) with a demonstrated emissions reduction potential from such operations (Hoffrichter et al. 2012). However, additional supportive policies will most likely be required (Strale 2016; Watson et al. 2019). Limiting emissions avoidance assessments for HSR modal substitution to account only for CO2 emissions ignores aviation’s non-CO2 effects (Section 10.5.2), and likely results in an under-representation of the climate benefits of HSR replacing flights.

HSR modal substitution can generate a contra-effect if the air traffic departure and arrival slots that become available as the result of the modal shift are simply reallocated to additional air services (Givoni and Banister 2006; Givoni and Dobruszkes 2013; Jiang and Zhang 2016; Cornet et al. 2018; Zhang et al. 2019a). Furthermore, HSR services have the potential to increase air traffic at a hub airport through improved networks but this effect can vary based on the distance of the HSR stations from airports (Jiang and Zhang 2014; Xia and Zhang 2016; Zhang et al. 2019b; Liu et al. 2019). Such rebound effects could be managed through policy interventions. For example, in 2021 the French government regulated that all airlines operating in France suspend domestic airline flights on routes if a direct rail alternative with a travel time of less than 2.5 hours is available. Other air travel demand reduction measures that have been proposed include regulations to ban frequent flyer reward schemes, mandates that all marketing of air travel declare flight emissions information to the prospective consumer (i.e., the carbon footprint of the nominated flight), the introduction of a progressive ‘Air Miles Levy’ as well as the inclusion of all taxes and duties that are presently exempt from air ticketing (Carmichael 2019). Moreover, China has the highest use of HSR in the world in part due to its network and competitive speeds and in part due to heavy regulation of the airline industry, in particular restrictions imposed on low-cost air carrier entry and subsidisation of HSR (Li et al. 2019). These air travel demand reduction strategies may induce shifts to other alternative modes in addition to stimulating HSR ridership.

Despite the risk of a rebound effect, and due to the probable reality of an incremental adoption of sustainable aviation fuel technology in the coming decades, the commencement of appropriate HSR services has the potential to provide, particularly in the short- to medium-term, additional means of aviation emissions mitigation.

The most recent projection from ICAO (prior to the COVID-19 pandemic) for international traffic (mid-range growth) is shown in Figure 10.11. This projection shows the different contributions of mitigation measures from two levels of improved technology, as well as improvements in air traffic management and infrastructure use. The projections indicate an increase in CO2 emissions by a factor of 2.2 in 2050 over 2020 levels for the most optimistic set of mitigation assumptions. The high/low traffic growth assumptions would indicate increases by factors of 2.8 and 1.1, respectively in 2050, over 2020 levels (again, for the most optimistic mitigation assumptions).

The International Energy Agency has published several long-term aviation scenarios since AR5 within a broader scope of energy projections. Their first set of aviation scenarios include a ‘reference technology scenario’, a ‘2°C Scenario’ and a ‘Beyond 2°C Scenario’. The scenarios are simplified in assuming a range of growth rates and technological/operational improvements (IEA 2017b). Mitigation measures brought about by policy and regulation are treated in a broad-brush manner, noting possible uses of taxes, carbon pricing, price and regulatory signals to promote innovation.

The IEA has more recently presented aviation scenarios to 2070 in their ‘Sustainable Development Scenario’ that assume some limited reduction in demand post-COVID-19, and potential technology improvements in addition to direct reductions in fossil kerosene usage from substitution of biofuels and synthetic fuels (IEA 2021b). There is much uncertainty in how aviation will recover from the COVID-19 pandemic but, in this scenario, air travel returns to 2019 levels in three years, and then continues to expand, driven by income. Government policies could dampen demand (12% lower by 2040 than the IEA ‘Stated Policies Scenario’, which envisages growth at 3.4% per year, which in turn is lower than ICAO at 4.3%). Mitigation takes place largely by fuel substitution with lower-carbon biofuels and synthetic fuels, with a smaller contribution from technology. Approximately 85% of the actual cumulative CO2 emissions (to 2070) are attributed to use of fuel at their lowest technology readiness level of ‘Prototype’, which is largely made up of biofuels and synthetic fuels, as shown in Figure 10.12. Details of the technological scenarios and the fuel availability/uptake assumptions are given in IEA (2021b), which also makes clear that the relevant policies are not currently in place to make any such scenario happen.

Within the Coupled Model Intercomparison Project Phase 6 emissions database, a range of aviation emissions scenarios for a range of Shared Socio-economic Pathway (SSP) scenarios are available (Figure 10.13). This Figure suggests that by 2050, direct emissions from aviation could be 1.5 to 6.5 (5–95th percentile) times higher than in the 2020 model year under the scenarios that exceed warming of 4°C during the 21st century with a likelihood of 50% or greater (C8). In the C1 (which limit warming to 1.5°C (>50%) during the 2st century with no or limited overshoot) and C2 (which return warming to 1.5°C (>50%) during the 2st century after a high overshoot) scenarios, aviation emissions could still be up to 2.5 times higher in 2050 than in the 2020 model year (95th percentile) but may need to decrease by 10% by 2050 (5th percentile).

The COVID-19 pandemic of 2020 has changed many activities and, consequentially, associated emissions quite dramatically (Le Quéré et al. 2018b; Friedlingstein et al. 2020; Liu et al. 2020c; UNEP 2020). Aviation was particularly affected, with a reduction in commercial flights in April 2020 of about 74% over 2019 levels, with some recovery over the following months, remaining at 42% lower as of October 2020 (Petchenik 2021). The industry is considering a range of potential recovery scenarios, with the International Air Transport Association (IATA) speculating that recovery to 2019 levels may take up until 2024 (Earley and Newman 2021) (Cross-Chapter Box 1 in Chapter 1). Others suggest, however, that the COVID-19 pandemic and increased costs as a result of feed-in quotas or carbon taxes could slow down the rate of growth of air travel demand, though global demand in 2050 would still grow 57%–187% between 2018 and 2050 (instead of 250% in a baseline recovery scenario) (Gössling et al. 2021a).

Under Article 2.2 of the Kyoto Protocol, Annex I countries were called to ‘…pursue limitation or reduction of emissions of GHGs not controlled by the Montreal Protocol from aviation and marine bunker fuels, working through the International Civil Aviation Organization and the International Maritime Organization, respectively. ’ The Paris Agreement is different, in that ICAO (and the IMO) are not named. As a result, the Paris Agreement, through the NDCs, seemingly covers CO2 emissions from domestic aviation (currently 35% of the global total from aviation) but does not cover emissions from international flights. A number of states and regions, including the UK, France, Sweden, and Norway, have declared their intentions to include international aviation in their net zero commitments, while the EU, New Zealand, California, and Denmark are considering doing the same (Committee on Climate Change 2019). The Paris Agreement describes temperature-based goals, such that it is unclear how emissions of GHGs from international aviation would be accounted for. Clearly, this is a less than ideal situation for clarity of governance of international GHG emissions from both aviation and shipping. At its 40th General Assembly (October 2019) the ICAO requested its Council to ‘…continue to explore the feasibility of a long-term global aspirational goal for international aviation, through conducting detailed studies assessing the attainability and impacts of any goals proposed, including the impact on growth as well as costs in all countries, especially developing countries, for the progress of the work to be presented to the 41st Session of the ICAO Assembly’. What form this goal will take is unclear until work is presented to the 41st Assembly (Autumn, 2022). It is likely , however, that new accountability and governance structures will be needed to support decarbonisation of the aviation sector.

Maritime transport volume has increased by 250% over the past 40 years, reaching an all-time high of 11 billion tonnes of transported goods in 2018 (UNCTAD 2019). This growth in transport volumes has resulted in continued growth in GHG emissions from the shipping sector, despite an improvement in the carbon intensity of ship operations, especially since 2014. The estimated total emissions from maritime transport can vary depending on data set and calculation method, but range over 600–1100 MtCO2 yr –1 over the past decade (Figure 10.14), corresponding to 2–3% of total anthropogenic emissions. The legend in Figure 10.14 refers to the following data sources: Endresen et al. (2003), Eyring et al. (2005), Dalsøren et al. (2009), DNV GL (DNV GL 2019), CAMS-GLOB-SHIP (Jalkanen et al. 2014; Granier et al. 2019), EDGAR (Crippa et al. 2019), Hoesly et al. (2018), Johansson et al. (2017), ICCT (Olmer et al. 2017), the IMO GHG Studies; IMO 2nd (Buhaug et al. 2009), IMO 3rd (Smith et al. 2014), IMO 4th-vessel and IMO 4th-voyage (Faber et al. 2020), and Kramel et al. (2021).

Like aviation, shipping is also a source of emissions of SLCFs as described in Section 10.5, including nitrogen oxides (NOx), sulphur oxides (SOx, such as SO2 and SO4), carbon monoxide (CO), black carbon, and non-methane volatile organic compounds (NMVOCs) (Szopa et al. 2021). Though SLCF have a shorter lifetime than the associated CO2 emissions, these short-lived forcers can have both a cooling effect (e.g., SOx) or a warming effect (e.g., ozone from NOx). The cooling from the SLCF from a pulse emission will decay rapidly and diminish after a couple of decades, while the warming from the long-lived substances lasts for centuries (Szopa et al. 2021).

Emissions of SLCF from shipping not only affect the climate, but also the environment, air quality, and human health. Maritime transport has been shown to be a major contributor to coastal air quality degradation (Zhao et al. 2013; Jalkanen et al. 2014; Viana et al. 2014; Goldsworthy and Goldsworthy 2015; Goldsworthy 2017). Sulphur emissions may contribute towards acidification of the ocean (Hassellöv et al. 2013). Furthermore, increases in sulphur deposition on the oceans have also been shown to increase the flux of CO2 from the oceans to the atmosphere (Hassellöv et al. 2013). To address the risks of SOx emissions from shipping, there is now a cap on the on the sulphur content permissible in marine fuels (IMO 2013). There is also significant uncertainty about the impacts of pollutants emitted from ships on the marine environment (Blasco et al. 2014).

Pollution control is implemented to varying degrees in the modelling of the SSP scenarios (Rao et al. 2017); for example, SSPs 1 and 5 assume that increasing concern for health and the environment result in more stringent air pollution policies than today (Szopa et al. 2021). There is a downward trend in SOx and NOx emissions from shipping in all the SSPs, in compliance with regulations. The SLCF emissions reduction efforts, within the maritime sector, are also contributing towards achieving the UN SDGs. In essence, while long-lived GHGs are important for long-term mitigation targets, accounting for short-lived climate forcers is important both for current and near-term forcing levels as well as broader air pollution and SDG implications.

Shipping in the Arctic is a topic of increasing interest. The reduction of Arctic summer sea ice increases the access to the northern sea routes (Smith and Stephenson 2013; Melia et al. 2016; Aksenov et al. 2017; Fox-Kemper et al. 2021). Literature and public discourse has sometimes portrayed this trend as positive (Zhang et al. 2016b), as it allows for shorter shipping routes, for example between Asia and Europe, with estimated travel time savings of 25–40% (Aksenov et al. 2017). However, the acceleration of Arctic cryosphere melt and reduced sea ice that enable Arctic shipping reduce surface albedo and amplify climate warming (Eyring et al. 2021). Furthermore, local air pollutants can play different roles in the Arctic. For example, black carbon emissions reduce albedo and absorb heat in air, on snow and ice (Browse et al. 2013; Kang et al. 2020; Messner 2020; Eyring et al. 2021). Finally, changing routing from Suez to the northern sea routes may reduce total emissions for a voyage, but also shifts emissions from low to high latitudes. Changing the location of the emissions adds complexity to the assessment of the climatic impacts of Arctic shipping, as the local conditions are different and the SLCF may have a different impact on clouds, precipitation, albedo and local environment (Dalsøren et al. 2013; Fuglestvedt et al. 2014; Marelle et al. 2016). Observations have shown that 5–25% of air pollution in the Arctic stems from shipping activity within the Arctic itself (Aliabadi et al. 2015). Emissions outside the Arctic can affect Arctic climate, and changes within the Arctic may have global climate impacts. Both modelling and observations have shown that aerosol emissions from shipping can have a significant effect on air pollution and shortwave radiative forcing (Ødemark et al. 2012; Peters et al. 2012; Dalsøren et al. 2013; Roiger et al. 2014; Righi et al. 2015; Marelle et al. 2016).

Increased Arctic shipping activity may also pose increased risks to local marine ecosystems and coastal communities from invasive species, underwater noise, and pollution (Halliday et al. 2017; IPCC 2019). Greater levels of Arctic maritime transport and tourism have political, as well as socio-economic, implications for trade, and nations and economies reliant on the traditional shipping corridors. There has been an increase in activity from cargo, tankers, supply, and fishing vessels in particular (Winther et al. 2014; Zhao et al. 2015). Projections indicate more navigable Arctic waters in the coming decades (Smith and Stephenson 2013; Melia et al. 2016) and continued increases in transport volumes through the northern sea routes (Corbett et al. 2010; Lasserre and Pelletier 2011; Winther et al. 2014). Emission patterns and quantities, however, are also likely to change with future regulations from IMO, and depend on technology developments, and activity levels which may depend upon geopolitics, commodity pricing, trade, natural resource extraction, insurance costs, taxes, and tourism demand (Johnston et al. 2017). The need to include indigenous peoples’ voices when shaping policies and governance of shipping activities in the high north is increasing (Dawson et al. 2020).

The Arctic climate and environment pose unique hazards and challenges with regard to safe and efficient shipping operations: low temperature challenges, implications for vessel design, evacuation and rescue systems, communications, oil spills, variable sea ice, and meteorological conditions (Buixadé Farré et al. 2014). To understand the total implications of shipping in the Arctic, including its climate impacts, a holistic view of synergies, trade-offs, and co-benefits is needed, with assessments of impacts on not only the physical climate, but also the local environment and ecosystems. To further ensure safe operations in the Arctic waters, close monitoring of activities may be valuable.

A range of vessel mitigation options for the international fleet exist and are presented in this section. A variety of feedstocks and energy carriers can be considered for shipping. As feedstocks, fuels from biomass (advanced biofuels), fuels produced from renewable electricity and CO2 capture from flue gas or the air (electro-, e-, or power-fuels), and fuels produced via thermochemical processes (solar fuels) can be considered. As energy carriers, synthetic fuels and the direct use of electricity (stored in batteries) are of relevance. The most prominent synthetic fuels discussed in the literature are hydrogen, ammonia, methane, methanol, and synthetic hydrocarbon diesel. Figure 10.15 shows the emissions reductions potential for alternative energy carriers that have been identified as having the highest potential to mitigate operational emissions from the sector (Chatzinikolaou and Ventikos 2014; Brynolf et al. 2014; Teeter and Cleary 2014; Traut et al. 2014; Lindstad et al. 2015; Psaraftis 2015; Seddiek 2015; Tillig et al. 2015; Winkel et al. 2016; DNV GL 2017; Bicer and Dincer 2018a; Biernacki et al. 2018; Bongartz et al. 2018; Gilbert et al. 2018; Hua et al. 2018; ITF 2018b; Singh et al. 2018; Balcombe et al. 2019; Hansson et al. 2019; Sharafian et al. 2019; Winebrake et al. 2019; Czermański et al. 2020; Faber et al. 2020; Hansson et al. 2020; Kim et al. 2020; Liu et al. 2020a; Nguyen et al. 2020; Perčić et al. 2020; Sadeghi et al. 2020; Seithe et al. 2020; Xing et al. 2020; Valente et al. 2021; Stolz et al. 2021).

Low-carbon hydrogen and ammonia are seen to have positive potential as a decarbonised shipping fuel. Hydrogen and ammonia, when produced from renewables or coupled to CCS as opposed to mainly by fossil fuels with high lifecycle emissions (Bhandari et al. 2014), may contribute to significant CO2-eq reductions of up to 70–80% compared to low-sulphur heavy fuel oil (Bicer and Dincer 2018b; Gilbert et al. 2018). These fuels have their own unique transport and storage challenges as ammonia requires a pilot fuel due to difficulty in combustion, and ammonia combustion could lead to elevated levels of NOx, N2O, or NH3 emissions depending on engine technology used (DNV GL 2020). There is a need for the further development of technology and procedures for safe storage and handling of fuels such as hydrogen and ammonia, both onboard and onshore, for faster uptake (Hoegh-Guldberg et al. 2019), but they remain an encouraging decarbonisation option for shipping in the next decade.

While methanol produced from fossil sources induces an emissions increase of +7.5% (+44%), e-methanol (via hydrogen from electrolysis based on renewable energy and carbon from direct air capture) reduces emissions by 80% (82%). In general, several synthetic fuels, such as synthetic diesel, methane, methanol, ethanol, and dimethyl ether could in principle be used for shipping (Horvath et al. 2018). The mitigation potential of these is fully dependent on the sourcing of the hydrogen and carbon required for their synthesis.

As noted in Section 10.3, LNG has been found to have a relatively limited mitigation potential and may not be viewed as a low-carbon alternative, but has a higher availability than other fuel options (Gilbert et al. 2018). Emissions reductions across the full fuel lifecycle are found in the order of 10%, with ranges reported from –30% (reduction) to +8% (increase), if switching from heavy fuel oil to LNG, as indicated in Figure 10.15 (Bengtsson et al. 2011). Regardless of the production pathway, the literature points to the risk of methane slip (emissions of unburnt methane especially at low engine loads and from transport to ports) from LNG-fuelled vessels, with no current regulation on emissions caps (Anderson et al. 2015; Ushakov et al. 2019; Peng et al. 2020). Leakage rates are a critical point for the total climate impact of LNG as a fuel, where high pressure engines remedy this more than low pressure ones. As discussed in Section 10.3, some consider LNG as a transition fuel, while some literature points to the risk of stranded assets due to the increasing decarbonisation regulation from IMO and the challenge of meeting IMO’s 2030 emissions reductions targets using this fuel.

In addition to fossil and e-fuels, advanced biofuels might play a role to provide the energy demand for future shipping. Biomass is presently used to produce alcohol fuels (such as ethanol and methanol), liquid biogas, or biodiesel that can be used for shipping and could reduce CO2 emissions from this segment. As explained in Box 10.2 and Chapter 7, the GHG footprint associated with biofuels is strongly dependent on the incurred land use and land-use change emissions. Advanced biofuels from processing cellulose rather than sugar are likely to be more attractive in terms of the quantities required but are not commercially available (Section 10.3). The estimates of emissions reductions from biofuels shown in Figure 10.15 rely on data from the Integrated Assessment Models – Energy Modelling Forum 33 (IAM EMF33), partial models assuming constant land cover (CLC), and partial models using natural growth (NRG). Box 10.2 and Section 10.4 include a more detailed description of the assumptions underlying these models and their estimates. The results based on IAM EMF33 and CLC suggests median mitigation potential of around 73% for advanced biofuels in shipping, while the NRG-based results suggest increased emissions from biofuels. The EMF33 and CLC results rely on modelling approaches compatible with the scenarios in the AR6 database (Chapter 6 and Box 7.7).

In addition to fuels, there are other measures that may aid the transition to low-carbon shipping. The amounts and speed of uptake of alternative low- or zero-carbon fuels in ports depend upon investments in infrastructure – including bunkering infrastructure, refinery readiness, reliable supply of the fuels, as well as sustainable production. The ship lifetime and age also play a role; retrofitting ships to accommodate engines and fuel systems for new fuel types may not be an option for older vessels. As such, operational efficiency becomes more important (Bullock et al. 2020). There is some potential to continue to improve the energy efficiency of vessels through operational changes (Traut et al. 2018), reducing the speed or ‘slow steaming’ (Bullock et al. 2020), and improved efficiency in port operations (Viktorelius and Lundh 2019; Poulsen and Sampson 2020). There is also a growing interest in onboard technologies for capturing carbon, with prototype ships underway showing 65–90% potential reduction in CO2 emissions (Luo and Wang 2017; Awoyomi et al. 2020; Japan Ship Technology Research Association et al. 2020). Challenges identified include CO2 capture efficiency (Zhou and Wang 2014), increased operating costs, and limited onboard power supply (Fang et al. 2019). Furthermore, designing CO2 storage tanks for transport to shore may pose a challenge, as the volume and weight of captured CO2 could be up to four times more than standard oil (Decarre et al. 2010).

Changes in design and engineering provide potential for reducing emissions from shipping through a range of measures, for example by optimising hull design and vessel shape, power and propulsion systems that include wind- or solar-assisted propulsion, and through improved operations of vessels and ports. Figure 10.15 shows that such measures may decrease emissions by 5–40%, though with a broad range in potential (Bouman et al. 2017). Nuclear propulsion could decrease emissions from individual vessels by 98%. Battery- or hybrid-electric ships have been identified as a means to reduce emissions in short-sea shipping such as ferries and inland waterways (Gagatsi et al. 2016), which may also importantly reduce near-shore SLCF pollution (Nguyen et al. 2020). Figure 10.15 shows that the median emissions from electric ships can be about 40% lower than equivalent fossil-based vessels but can vary widely. The wide reduction potential of battery electric propulsion is due to different assumptions about the CO2 intensity of the electricity used and the levels of CO2 footprints associated with battery production.

Although projections indicate continued increase in freight demand in the future, demand-side reductions could contribute to mitigation. The development of autonomous systems may play a role (Colling and Hekkenberg 2020; Liu et al. 2021) while 3-D printing can reduce all forms of freight as parts and products can be printed instead of shipped (UNCTAD 2018). As more than 40% of transported freight is fossil fuels, a lessened demand for such products in low-emissions scenarios should contribute to reducing the overall maritime transport needs and hence emissions in the future (Sharmina et al. 2017). An increase in alternative fuels, on the other hand, may increase freight demand (Mander et al. 2012). Potentials for demand-side reduction in shipping emissions may arise from improving processes around logistics and packaging, and further taxes and charges could serve as leverage for reducing demand and emissions.

The coming decade is projected to be costly for the shipping sector, as it is preparing to meet the 2030 and 2050 emissions reduction targets set by the IMO (UNCTAD 2018). With enough investments, incentives, and regulation, substantial reductions of CO2 emissions from shipping could be achieved through alternative energy carriers. The literature suggests that their cost could be manyfold higher than for conventional fuels, which in itself could reduce demand for shipping, and hence its emissions, but could make the transition difficult. R&D may help reduce these costs. The literature points to the need for developing technology roadmaps for enabling the maritime transport sector to get on to pathways for decarbonisation early enough to reach global goals (Kuramochi et al. 2018). Accounting for the full lifecycle emissions of the vessels and the fuels is required to meet the overall long-term objectives of cutting GHG and SLCF emissions. The urgency of implementing measures for reducing emissions is considered to be high, considering the lifetime of vessels is typically 20 years, if not more.

Regulatory frameworks for the shipping sector have been developed over time and will continue to be through bodies such as the IMO, which was established by the UN to manage international shipping. The IMO strategy involves a 50% reduction in GHG emissions from international shipping by 2050 compared to 2008 (IMO 2018). The strategy includes a reduction in carbon intensity of international shipping by at least 40% by 2030, and 70% by 2050, compared to 2008. IMO furthermore aims for the sectoral phase-out of GHG emissions as soon as possible this century.

In 2020, the IMO approved the short-term goal-based measure to reduce the carbon intensity of existing international vessels. This measure addresses both technical and operational strategies. The operational element is represented by a Carbon Intensity Indicator (CII), and the technical element is represented by the Energy Efficiency Existing Ship Index (EEXI), which will apply to ships from 2023. The EEXI builds upon the Energy Efficiency Design Index (EEDI), which is a legally-binding mitigation regulation for newbuild ships, established as a series of baselines for the amount of fuel ships may burn for a particular cargo-carrying capacity. The EEDI differs per ship segment. For example, ships built in 2022 and beyond should be 50% more energy efficient than those built in 2013. This legislation aims to reduce GHG emissions in particular. Energy efficiency may be improved by several of the mitigation options outlined above. The Ship Energy Efficiency Management Plan (SEEMP) is seen as the international governance instrument to improve energy efficiency and hence emissions from ships. SEEMP is a measure to enable changes to operational measures and retrofits (see Johnson et al. 2013). The combination of EEXI, EEDI, and SEEMP may reduce emissions by 23% by 2030 compared to a ‘no policy’ scenario (Sims et al. 2014). With regards to accountability, it is mandatory for ships greater than or equal to 5000 gross tonnage to collect fuel consumption data, as well as specified data. Such as for transport work. Similarly, the EU Monitoring, Reporting and Verification Regulation requires mandatory reporting of a vessel’s fuel consumption when operating in European waters.

Policy choices may enable or hinder changes, and gaps in governance structures may, to some degree, hinder the objectives of mechanisms like SEEMP to improve energy efficiency and emissions. Policies may be developed to incentivise investments in necessary changes to the global fleet and related infrastructures. The literature argues that regulations and incentives that motivate mitigation through speed optimisation, ship efficiency improvements, and retrofits with lower-carbon technologies at a sub-global scale may contribute to immediate reductions in CO2 emissions from the sector (Bows-Larkin 2015). The role of the financial sector, through initiatives such as the Poseidon Principle, which limit lending to companies that fail to uphold environmental standards, could also become increasingly important (Sumaila et al. 2021).

It has been proposed to make shipping corporations accountable for their emissions by making it mandatory to disclose their vessels’ emissions reductions (Rahim et al. 2016). Market-based mechanisms may increasingly encourage ship operators to comply with IMO GHG regulations. Development of policies such as carbon pricing or taxation to enable a business case for adopting low-carbon fuels could be a near-term priority for acceleration of transformation of the sector (Hoegh-Guldberg et al. 2019). The EU is considering including shipping in its carbon trading system, with the details still to be agreed upon but expected to come into force in 2023, along with the CII. The proposition is that shipowners who conduct voyages within Europe, or start or end at an EU port, will have to pay for carbon permits to cover the CO2 emitted by their vessel.

Regulations exist also to limit emissions of air pollution from shipping with the aim to improve environment and health impacts from shipping in ports and coastal communities. In sulphur emission control areas (SECAs), the maximum permissible sulphur content in marine fuels is 0.10% mass/mass. These are further tightened by the IMO legislation on reducing marine fuel sulphur content to a maximum of 0.5% in 2020 outside SECAs, compared to 3.5% permissible since 2012 (MARPOL Convention). The MARPOL Annex VI also limits the emissions of ozone-depleting substances and ozone precursors, NOx, and volatile organic compounds from tankers (Mertens et al. 2018). The implementation of the emission control areas have been shown to reduce the impacts on health and the environment (Viana et al. 2015).

While there are many governance and regulatory initiatives that help reduce emissions from the shipping sector, few are transformative on their own, unless zero-carbon fuels can become available at a reasonable cost as suggested in Sections 10.3 and below.

Figure 10.16 shows CO2 emissions from shipping in scenarios from the AR6 database and the Fourth GHG study by the IMO (Faber et al. 2020). Panel (a) shows that CO2 emissions from shipping go down by 33–70% (5–95th% percentile) by 2050 in the C1 and C2 scenarios, which limit warming to 1.5°C (>50%) during the 21st century with no or limited overshoot or return warming to 1.5°C (>50%) during the 21st century after a high overshoot. By 2080, median values for the same set of scenarios reach net zero CO2 emissions. IAMs often do not report emissions pathways for shipping transport and the sector is underrepresented in most IAMs (Esmeijer et al. 2020). Hence pathways established outside IAMs can be different for the sector. Indeed, the IMO projections for growth in transport demand (Faber et al. 2020) indicate increases of 40–100% by 2050 for the global fleet. Faber and et al. (2020), at the same time predict reductions in trade for fossil fuels dependent on decarbonisation trajectories. The energy efficiency improvements of the vessels in these scenarios are typically of 20–30%. This offsets some of the increases from higher demand in the future scenarios. Fuels assessed by the Fourth IMO GHG study were limited to heavy fuel oil, marine gasoil, LNG, and methanol, with a fuels mix ranging from 91–98% conventional fuel use and a small remainder of alternative fuels (primarily LNG and some methanol). Panel (b) shows average fleetwide emissions of CO2 based on these aggregate growth and emissions trajectories from the IMO scenarios. In these scenarios, CO2 emissions from shipping remain stable or grow compared to 2020 modelled levels. These results contrast with the low emissions trajectories in the C1–C2 bin in panel (a). It seems evident that the scenarios in the AR6 database explore a broader solutions space for the sector than the Fourth GHG study by the IMO. However, the 1.5°C–2°C warming goal has led to an IMO 2050 target of 40% reduction in carbon intensity by 2030, which would require emissions reduction efforts to begin immediately. Results from global models suggest the solutions space for deep emissions reductions in shipping is available.

Combinations of measures are likely to be needed for transformative transitioning of the shipping sector to a low-carbon future, particularly if an expected increase in demand for shipping services is realised (Smith et al. 2014; Faber et al. 2020). Both GHG and SLCF emissions decrease significantly in SSP1-1.9, where mitigation is achieved in the most sustainable way (Rao et al. 2017). Conversely, there are no emissions reductions in the scenarios presented by the IMO Fourth GHG study, even though these scenarios incorporate some efficiency improvements and a slight increase in the use of LNG.

Options outlined in this chapter suggest a combination of policies to reduce demand, increase investments by private actors and governments, and develop the technology readiness level of alternative fuels and related infrastructure (especially synthetic fuels). Some literature suggests that battery electric-powered short-distance sea shipping could yield emissions reductions given access to low-carbon electricity. For deep sea shipping, advanced biofuels, hydrogen, ammonia, and synthetic fuels hold potential for significant emissions reductions, depending on GHG characteristics of the fuel chain and resource base. Other options, such as optimisation of speed and hull design and wind-assisted ships, could also combine to make significant contributions by 2050 to further bring emissions down. In total a suite of mitigation options exists or is on the horizon for the maritime sector.

This section reviews the results of three types of models that systemically combine options to assess different approaches to generating decarbonisation pathways for the transport system: (i) integrated assessment models (IAM); (ii) global transport energy models (GTEM); and (iii) national transport-energy models (NTEM) (Edelenbosch et al. 2017; Yeh et al. 2017). Common assumptions across the three model types include trajectories of socioeconomic development, technological development, resource availability, policy, and behavioural change. The key differences underlying these models are their depth of technological and behavioural detail versus scope in terms of sectoral and regional coverage. In very general terms, the narrower the scope in terms of sectors and regions, the more depth on spatial, technological, and behavioural detail. A large set of scenarios from these models were collected in a joint effort led by Chapter 3 and supported by Chapter 10 and others. The outcomes from over 100 models have been analysed for this chapter with the methodologies set out in Annex III for the whole report.

GHG emissions from transport are a function of travel demand, travel mode, transport technology, GHG intensity of fuels, and energy efficiency. These drivers can be organised around a group of levers that can advance the decarbonisation of the transport system. The levers thus include reducing travel activity, increasing use of lower-carbon modes, and reducing modal energy intensity and fuel carbon content. This section explores each lever’s contributions to the decarbonisation of the transport sector by reviewing the results from the three model types IAM, GTEM, and NTEM.

IAMs integrate factors from other sectors that interact with the transport system endogenously, such as fuel availability and costs. IAMs minimise mitigation costs to achieve a temperature goal across all sectors of the economy over a long time horizon (typically to 2100). IAMs typically capture mitigation options for energy and carbon intensity changes with greater technology/fuel details and endogeneity linked to the other sectors. In the scenarios with very large-scale electrification of the transport sector, the coupling with the other sectors in fuel production, storage, and utilisation becomes more important. G-/NTEMs and related regional transport sectoral models have more details on transport demand, technology, behaviours, and policies than IAMs, but treat the interactions with the other sectors exogenously, potentially missing some critical interactions, such as the fuel prices and carbon intensity of electricity. National models have detailed representation of national policies related to transport and energy, sometimes with greater spatial resolution. Compared with IAMs, G-/NTEMs typically have greater detailed representation to explore mitigation options along the activity and mode dimensions where spatial, cultural, and behavioural details can be more explicitly represented. Section 5 in Annex III provides more details about these types of models. Scenarios for shipping and aviation are handled in more detail in sections 10.5 and 10.6, respectively.

This section applies the following categorisation of scenarios (see Table 3.1 for more details):

•C1 (scenarios that limit warming to 1.5°C (>50%) during the 21st century with no or limited overshoot)

•C2 (scenarios that return warming to 1.5°C (>50%) during the 21st century after a high overshoot)

•C3 (scenarios that limit warming to 2°C (>67%) throughout the 21st century)

•C4 (scenarios that limit warming to 2°C (>50%) throughout the 21st century)

•C5 (scenarios that limit warming to 2.5°C (>50%) throughout the 21st century)

•C6 (scenarios that limit warming to 3°C (>50%) throughout the 21st century)

•C7 (scenarios that limit warming to 4°C (>50%) throughout the 21st century)

•C8 (scenarios that exceed warming of 4°C (≥50%) during the 21st century)

A large share of the scenarios was developed prior to 2020. Results from such scenarios are indexed to a modelled (non-COVID) year 2020, referred to as 2020Mod.

In 2018, transport emitted 8.5 GtCO2-eq, reaching a near doubling from 1990 levels after two decades of 2% per year emissions growth (Section 10.1). Assessing future trajectories, Figure 10.17 provides an overview of direct CO2 emissions estimates from the transport sector across IAMs (colour bars) and selected global transport models (grey bars). The results from the IAMs are grouped in bins by temperature goal. Global transport energy models are grouped into reference and policy bins, since the transport sector cannot by itself achieve fixed global temperature goals. The policy scenarios in GTEMs and NTEMs cover a wide range of ‘non-reference’ scenarios, which include, for example, assumptions based on the ‘fair share action’ principles. In these scenarios, transport emissions reach reductions consistent with the overall emissions trajectories aligning with warming levels of 2°C. These scenarios may also consider strengthening existing transport policies, such as increasing fuel economy standards or large-scale deployments of electric vehicles. In most cases, these Policy scenarios are not necessarily in line with the temperature goals explored by the IAMs.

According to the collection of simulations from the IAM and GTEM models shown in Figure 10.17, global transport emissions could grow up to 2–47% (5–95th percentile) by 2030 and –6–130% by 2050 under the C7 scenarios that limit warming to 4°C (>50%) throughout the 21st century and C8 scenarios that exceed 4°C (≥50%) during the 21st century. Population and GDP growth and the secondary effects, including higher travel service demand per capita and increased freight activities per GDP, drive the growth in emissions in these scenarios (Section 10.7.3). Though transport efficiencies (energy use per pkm travelled and per tkm of goods delivered) are expected to continue to improve in line with the historical trends (Section 10.7.4), total transport emissions would grow due to roughly constant carbon intensity (Section 10.7.5) under the C7 and C8 scenarios that limit warming to 4°C (>50%) throughout the 21st century or exceed 4°C (≥50%) during the 21st century. In these scenarios, Significant increases in emissions (>150% for the medium values by 2050) would come from Asia and Pacific, the Middle East, and Africa. Compared to estimated 2020 levels, in 2050 Developed Countries would have median 25% decrease in transport emissions in the C7 scenarios that limit warming to 4°C (>50%) throughout the 21st century or median 15% increase in transport emissions in the C8 scenarios that exceed warming of 4°C (≥50%) during the 21st century.

To meet temperature goals, by 2050 global transport emissions would need to decrease by 17% (+67% to –23% for the 5–95th percentile) below 2020Mod levels in the scenarios that limit warming to 2°C (>67%), 2°C (>50%) and 2.5 °C (>50%) throughout the 21st century (C3-C5 scenarios – orange bars), and 47% (14–80% for the 5–95th percentile) in the scenarios that limit warming to 1.5°C (>50%) during the 21st century with no or limited overshoot or return to 1.5°C (>50%) during the 21st century after high overshoot during the 21st century (C1–C2 scenarios – green bars). However, transport-related emission reductions may not happen uniformly across regions. For example, transport emissions from the Developed Countries and Eastern Europe and West Central Asia would decrease from 2020 levels by 2050 across all C1–C2 scenarios, but could increase in Africa, Asia and Pacific, Latin America and Caribbean, and the Middle East, in some of these scenarios. In particular, the median transport emissions in India and Africa could increase by 2050 in C1–C2 scenarios, while the 95th percentile emissions in Asia and Pacific, Latin America and Caribbean, and the Middle East, could be higher in 2050 than in 2020.

The Reference scenario emission pathways from GTEMs described in Figure 10.17 have similar ranges to C7–C8 scenario groups in 2050. The Policy scenarios are roughly in line with C6–C7 scenarios for the world region. The results suggest that the majority of the Policy scenarios examined by the GTEMS reviewed here are in the range of the C3–C6 scenarios examined by the IAMs (Gota et al. 2016; IEA 2017b; Yeh et al. 2017; Fisch-Romito and Guivarch 2019). The NDCs in the transport sector include a mix of measures targeting efficiency improvements of vehicles and trucks; improving public transit services; decarbonising fuels with alternative fuels and technologies including biofuels, fossil- or bio-based natural gas, and electrification; intelligent transport systems; and vehicle restrictions (Gota et al. 2016). Because of the long lag-time for technology turnover, these measures are not expected to change 2030 emissions significantly. However, they could have greater impacts on 2050 emissions.

Several GTEMs not included in AR6 scenario database have examined ambitious CO2 mitigation scenarios. For example, a meta-analysis of scenarios suggests that global transport emissions consistent with warming levels of 2°C, would peak in 2020 at around 7–8 GtCO2 and decrease to 2.5–9.2 Gt for 2°C, with an average of 5.4 Gt by 2050 (Gota et al. 2019). For comparison, the IEA’s Sustainable Development Scenario suggests global transport emissions decrease to 3.3 Gt (or 55% reduction from 2020 level) by 2050 (IEA 2021f). The latest IEANet Zero by 2050 report proposes transport emissions to be close to zero by 2050 (IEA 2021e). The latter is lower than the interquartile ranges of the C1 group of scenarios from the AR6 database analysed here.

Low-carbon scenarios are also available from national models (Latin America, Brazil, Canada, China, France, Germany, Indonesia, India, Italy, Japan, Mexico, South Africa, UK, US) with a good representation of the transport sector. The low-carbon scenarios are either defined with respect to a global climate stabilisation level of, for example, 2°C/1.5°C Scenario (Dhar et al. 2018), or a CO2 target that is more stringent than what has been considered in the NDCs, such as the net-zero emissions pathways (Bataille et al. 2020; IEA 2021e). These studies have generally used bottom-up models (see Annex III) for the analysis, but in some cases, they are run by national teams using global models (e.g., the Global Change Assessment Model (GCAM) for China and India). National studies show that transport CO2 emissions could decline significantly in low-carbon scenarios in all the developed countries reviewed (Bataille et al. 2015; Kainuma et al. 2015; Hillebrandt et al. 2015; Mathy et al. 2015; Pye et al. 2015; Virdis et al. 2015; Williams et al. 2015; Zhang et al. 2016a) in 2050 from the emissions in 2010 and reductions could vary from 65% to 95%. However, in developing countries reviewed (Di Sbroiavacca et al. 2014; Altieri et al. 2015; Buira and Tovilla 2015; Rovere et al. 2015; Shukla et al. 2015; Siagian et al. 2015; Teng et al. 2015; Dhar et al. 2018), emissions could increase in 2050 in the range of 35% to 83% relative to 2010 levels. Transport CO2 emissions per capita in the developing countries were much lower in 2010 (varying from 0.15 to 1.39 tCO2 per capita) relative to developed countries (varying from 1.76 to 5.95 tCO2 per capita). However, results from national modelling efforts suggest that, by 2050, the CO2 emissions per capita in developed countries (varying from 0.19 to 1.04 tCO2 per capita) could be much lower than in developing countries (varying from 0.21 to 1.7 tCO2 per capita).

The transport scenario literature’s mean outcomes suggest that the transport sector may take a less steep emissions reduction trajectory than the cross-sectoral average and still be consistent with the 2°C goal. For example, most of the scenarios that limit or return warming to 1.5°C (>50%) during the 21st century (C1–C2) reach zero emissions by 2060, whereas transport sector emissions are estimated in the range of 20% of the 2020Mod level (4–65% for the 10th to 90th percentiles) by 2100. This finding is in line with perspectives in the literature suggesting that transport is one of the most difficult sectors to decarbonise (Davis et al. 2018). There is, however, quite a spread in the results for 2050. Since temperature warming levels relate to global emissions from all sectors, modelling results from IAMs tend to suggest that in the short and medium term, there might be lower cost mitigation options outside the transport sector. On the other hand, compared with GTEMs/NTEMs, some IAMs may have limited mitigation options available, including technology, behavioural changes, and policy tools especially for aviation and shipping. The models therefore rely on other sectors and/or negative emissions elsewhere to achieve the overall desired warming levels. This potential shortcoming should be kept in mind when interpreting the sectoral results from IAMs.

Growth in passenger and freight travel demand is strongly dependent on population growth and GDP. In 2015, transport activities were estimated at around 35–50 trillion pkm, or 5,000–7,000 pkm per person per year, with significant variations among studies (IEA 2017b; ITF 2019). The number of passenger cars in use has grown 45% globally between 2005–2015, with the most significant growth occurring in the developing countries of Asia and the Middle East (119%), Africa (79%), and South and Central America (80%), while growth in Europe and North America is the slowest (21% and 4% respectively) (IOMVM 2021). On the other hand, car ownership levels in terms of vehicles per 1000 people in 2015 were low in developing countries of Asia and the Middle East (141), Africa (42), South and Central America (176), while in Europe and North America they are relatively high (581 and 670 respectively) (IOMVM 2021). The growth rate in commercial vehicles (freight and passenger) was 41% between 2005 and 2015, with a somewhat more even growth across developed and developing countries (IOMVM 2021).

Figure 10.18 shows activity trajectories for both freight and passenger transport based on the AR6 database for IAMs. According to demand projections from the IAMs, global passenger and freight transport demand could increase relative to a modelled year 2020 across temperature goals. The median transport demand from IAMs for all the scenarios in line with warming levels below 2.5°C (C1–C5) suggests that global passenger transport demand could grow by 1.14–1.3 times in 2030 and by 1.5–1.8 times in 2050 (1.27–2.33 for the 5–95th percentile across C1–C5 scenarios) relative to modelled 2020 level. Developed regions including North America and Europe exhibit lower growth in passenger demand in 2050 compared to developing countries across all the scenarios. In 2030, most of the global passenger demand growth happens in Africa (44% growth relative to 2020), and Asia and Pacific (57% growth in China and 59% growth in India relative to 2020) in the scenarios that limit warming below 2.5°C (>50%) throughout the 21st century (C5). These regions start from a low level of per capita demand. For example, in India, demand may grow by 84%. However, the per capita demand in 2010 was under 7000 km per person per year (Dhar and Shukla 2015). Similarly, in China, demand may grow by 52%, starting from per capita demand of 8000 km per person per year in 2010 (Pan et al. 2018). The per capita passenger demand in these regions was lower than in developed countries in 2010, but it converges towards the per capita passenger transport demand of advanced economies in less stringent climate scenarios (C6–C7). Demand for passenger travel would grow at a slower rate in the stricter temperature stabilisation scenarios (<2.5°C and 1.5°C scenarios, C1–C5) compared to the scenarios with higher warming levels (C7–C8). The median global passenger demand in the scenarios that limit or return warming to 1.5°C during the 21st century (C1–C2) is 27% lower in 2050 relative to C8.

Due to limited data availability, globally consistent freight data is difficult to obtain. In 2015, global freight demand was estimated to be 108 trillion tkm, most of which was transported by sea (ITF 2019). The growth rates of freight service demand vary dramatically among different regions: over the 1975–2015 period, road freight activity in India increased more than 9-fold, 30-fold in China, and 2.5-fold in the US (Mulholland et al. 2018). Global freight demand continues to grow but at a slower rate compared to passenger demand across all the scenarios in 2050 compared to modelled 2020 values. Global median freight demand could increase by 1.17–1.28 times in 2030 and 1.18–1.7 times in 2050 in all the scenarios with warming below 2.5°C (C1–C5). Like passenger transport, the models suggest that a large share of growth occurs in Africa and Asian regions (59% growth in India and 50% growth in China in 2030 relative to a modelled year 2020) in the C5 scenarios that limit warming below 2.5°C (>50%) throughout the 21st century. Global median freight demand grows more slowly in the stringent temperature stabilisation scenarios, and is 40% and 22% lower in 2050 in the scenarios that limit or return warming to 1.5°C (>50%) during the 21st century (C1–C2) and below 2.5°C scenarios (C3–C4), respectively, compared to scenarios with warming levels of above 4°C (C8).

GTEMs show broad ranges for future travel demand, particularly for the freight sector. These results show more dependency on models than on baseline or policy scenarios. According to ITF Transport Outlook (ITF 2019), global passenger transport and freight demand could more than double by 2050 in a business-as-usual scenario. Mulholland et al. (2018) suggest the freight sector could grow 2.4-fold over 2015–2050 in the reference scenario, with the majority of growth attributable to developing countries. The IEA suggests a more modest increase in passenger transport, from 51 trillion pkm in 2014 to 110 trillion pkm in 2060, in a reference scenario without climate policies and a climate scenario that would limit emissions below 2°C. The demand for land-based freight transport in 2060 is, however, slightly lower in the climate scenario (116 trillion tkm) compared to the reference scenario (130 trillion tkm) (IEA 2017b). The ITF, however, suggests that ambitious decarbonisation policies could reduce global demand for passenger transport by 13–20% in 2050, compared to the business-as-usual scenario (ITF 2019; ITF 2021). The reduction in vehicle travel through shared mobility could reduce emissions from urban passenger transport by 30% compared to the business-as-usual scenario. Others suggest that reductions larger than 25%, on average, for both passenger and freight in 2030 and 2050 may be needed to achieve very low carbon emissions pathways (Fisch-Romito and Guivarch 2019). In the absence of large-scale carbon dioxide removal, few global studies highlight the need for significant demand reduction in critical sectors (aviation, shipping and road freight) in well below 2°C scenarios (van Vuuren et al. 2018; Grant et al. 2021; Sharmina et al. 2021).

Many models find small differences in passenger transport demand across temperature goals because IAM models rely on historical relationships between population, GDP, and demand for services to estimate future demand. This assumption poses a limitation to the modelling efforts, as mitigation efforts would likely increase travel costs that could result in lower transport demand (Zhang et al. 2018). In most models, demand is typically an exogenous input. These models often assume mode shifts of activities from the most carbon-intensive modes (driving and flying for passenger travel and trucking for freight) to less carbon-intensive modes (public transit and passenger rail, and freight rail) to reduce emissions.

Traditionally there is a disconnection between IAM models and bottom-up sectoral or city-based models due to the different scale (both spatial and temporal) and focus (climate mitigation vs urban pollution, safety (Creutzig 2016)). The proliferation of shared and on-demand mobility solutions is leading to rebound effects for travel demand (Chen and Kockelman 2016; Coulombel et al. 2019) and this is a new challenge for modelling. Some IAM studies have recently begun to explore demand-side solutions for reducing transport demand to achieve very low-carbon scenarios through a combination of culture and low-carbon lifestyle (Creutzig et al. 2018; van Vuuren et al. 2018); urban development (Creutzig et al. 2015a); increased vehicle occupancy (Grubler et al. 2018); improved logistics and streamlined supply chains for the freight sector (Mulholland et al. 2018); and disruptive low-carbon innovation, described as technological and business model innovations offering ‘novel value propositions to consumers and which can reduce GHG emissions if adopted at scale’ (Wilson et al. 2019). In the literature from national models, demand has been differentiated between conventional and sustainable development scenarios through narratives built around policies, projects, and programmes envisaged at the national level (Dhar and Shukla 2015; Shukla et al. 2015) and price elasticities of travel demand (Dhar et al. 2018). However, a greater understanding of the mechanisms underlying energy-relevant decisions and behaviours (Brosch et al. 2016), and the motivations for sustainable behaviour (Steg et al. 2015), are critically needed to realise these solutions.

Overall, passenger and freight activity are likely to continue to grow rapidly under the C7 (>3.0°C) scenarios, but most growth would occur in developing countries. Most models treat travel demand exogenously following the growth of population and GDP, but they have limited representation of responses to price changes, policy incentives, behavioural shifts, nor innovative mobility solutions that can be expected to occur in more stringent mitigation scenarios. Chapter 5 provides a more detailed discussion of the opportunities for demand changes that may result from social and behavioural interventions.

Globally over the last century, shares of faster transport modes have generally increased with increasing passenger travel demand (Schäfer 2017; Schafer and Victor 2000). For short- to medium-distance travel, private cars have displaced public transit, particularly in OECD countries, due to a variety of factors, including faster travel times in many circumstances (Liao et al. 2020); consumers increasingly valuing time and convenience with GDP growth; and broader transport policies, such as provision of road versus public transit infrastructure (Mattioli et al. 2020). For long-distance travel, travel via aviation for leisure and business has increased (Lee et al. 2021). These trends do not hold in all countries and cities, as many now have rail transit that is faster than driving (Newman et al. 2015). For instance, public transport demand rose from 1990 through to 2016 in France, Denmark, and Finland (eurostat 2019). In general, smaller and denser countries and cities with higher or increasing urbanisation rates tend to have greater success in increasing public transport share. However, other factors, like privatisation of public transit (Bayliss and Mattioli 2018) and urban form (ITF 2021), also play a role. Different transport modes can provide passenger and freight services, affecting the emissions trajectories for the sector.

Figure 10.19 shows activity trajectories for freight and passenger transport through 2100 relative to a modelled year 2020 across different modes, based on the AR6 database for IAMs and global transport models. Globally, climate scenarios from IAMs, and policy and reference scenarios from global transport models, indicate increasing demand for freight and passenger transport via most modes through 2100 (Yeh et al. 2017; Mulholland et al. 2018; Zhang et al. 2018; Khalili et al. 2019). Road passenger transport exhibits a similar increase (roughly tripling) through 2100 across scenarios. For road passenger transport, scenarios that limit or return warming to 1.5°C during the 21st century (C1–C2) have a smaller increase from modelled 2020 levels (median increase of 2.4 times modelled 2020 levels) than do scenarios with higher warming levels (C3–C8) (median increase of 2.7–2.8 times modelled 2020 levels). There are similar patterns for passenger road transport via light-duty vehicle, for which median increases from modelled 2020 levels are smaller for C1–C2 (3 times larger) than for C3–C5 (3.1 times larger) or C6–C7 (3.2 times larger). Passenger transport via aviation exhibits a 2.2 times median increase relative to modelled 2020 levels under C1–C2 and C3–C5 scenarios but exhibits a 6.2 times increase under C6–C8. The only passenger travel mode that exhibits a decline in its median value through 2100 according to IAMs is walking/bicycling, in C3–C5 and C6–C8 scenarios. However, in C1–C2 scenarios, walking/bicycling increases by 1.4 times relative to modelled 2020 levels. At the 5th percentile of IAM solutions (lower edge of bands in Figure 10.19), buses and walking/bicycling for passenger travel both exhibit significant declines.

For freight, Figure 10.19 shows that the largest growth occurs in transport via road (Mulholland et al. 2018). By 2100, global transport models suggest a roughly four-fold increase in median-heavy-duty trucking levels relative to modelled 2020 levels, while IAMs suggest a two- to four-fold increase in freight transport by road by 2100. Notably, the 95th percentile of IAM solutions see road transport by up to 4.7 times through 2100 relative to modelled 2020 levels, regardless of warming level. Other freight transport modes – aviation, international shipping, navigation, and railways – exhibit less growth than road transport. In scenarios that limit or return warming to 1.5°C (>50%) during the 21st century (C1–C2), navigation and rail transport remain largely unchanged and international shipping roughly doubles by 2100. Scenarios with higher warming (i.e., moving from C1–C2 to C6–C8) generally lead to more freight by rail and less freight by international shipping.

Relative to global trajectories, upper-income regions – including North America, Europe, and the Pacific OECD – generally see less growth in passenger road via light-duty vehicle and passenger aviation, given more saturated demand for both. Other regions like China exhibit similar modal trends as the global average, whereas regions such as the African continent and Indian subcontinent exhibit significantly larger shifts, proportionally, in modal transport than the globe. In particular, the African continent represents the starkest departure from global results. Freight and passenger transport modes exhibit significantly greater growth across Africa than globally in all available scenarios. Across Africa, median freight and passenger transport via road from IAMs increases by 5 to 16 times and 4 to 28 times, respectively, across warming levels by 2100 relative to modelled 2020 levels. Even C1 has considerable growth in Africa via both modes (3 to 16 times increase for freight and 4 to 29 times increase for passenger travel at 5th and 95th percentiles of IAM solutions by 2100).

As noted in Section 10.2, commonly explored mitigation options related to mode change include a shift to public transit, shared mobility, and demand reductions through various means, including improved urban form, teleconferences that replace passenger travel (Creutzig et al. 2018; Grubler et al. 2018; Wilson et al. 2019), improved logistics efficiency, green logistics, and streamlined supply chains for the freight sector (Mulholland et al. 2018). NDCs often prioritise options like bus improvements and enhanced mobility that yield pollution, congestion, and urban development co-benefits, especially in medium- and lower-income countries (Fulton et al. 2017). Conversely, high-income countries, most of which have saturated and entrenched private vehicle ownership, typically focus more on technology options, such as electrification and fuel efficiency standards (Gota et al. 2016). Available IAM and regional models are limited in their ability to represent modal shift strategies. As a result, mode shifts alone do not differentiate climate scenarios. While this lack of representation is a limitation of the models, it is unlikely that such interventions would completely negate the increases in demand the models suggest. Therefore, transport via light-duty vehicle and aviation, freight transport via road, and other modes will likely continue to increase through to the end of the century. Consequently, fuel and carbon efficiency and fuel energy and technology will probably play crucial roles in differentiating climate scenarios, as discussed in the following sub-sections.

This section explores what vehicle energy efficiencies and fuel carbon intensity trajectories, from the data available in the AR6 database from IAMs and GTEMs, could be compatible with different temperature goals. Figure 10.20 shows passenger and freight energy intensity, and fuel carbon intensity, indexed relative to 2020Mod values. The top panel shows passenger energy intensity across all modes. LDVs constitute a major share of this segment. Yeh et al. (2017) report 2.5–2.75 MJ vkm –1 in 2020 across models for the LDV segment, which is very close to the IEA estimate of 2.5 MJ vkm –1 for the global average fuel consumption for LDVs in 2017 (IEA 2020d). For reference, these numbers correspond to 1.6–1.7 MJ pkm –1 for an occupancy rate of 1.5. The following results of the AR6 database are conditional on the corresponding reductions in fuel carbon intensity. Figure 10.20 shows that the scenarios suggest that passenger transport’s energy intensity drops to between 10–23% (interquartile ranges across C1–C4) in 2030 for scenarios in line with warming levels below 2°C. In 2050, the medians across the group of scenarios that limit or return warming to 1.5°C (>50%) during the 21st century (C1–C2), and scenarios that limit warming to 2°C (>67% or >50%) throughout the 2st century (C3–C4) suggest energy intensity reductions of 51% and 45–46% respectively. These values correspond to annual average energy efficiency improvement rates of 2.3–2.4% and 2.0–2.1%, respectively, from 2020 to 2050. For reference, the IEA reports an annual energy efficiency improvement rate of 1.85% per year in 2005–16 (IEA 2020d). In contrast, the results from GTEMs suggest lower energy efficiency improvement, with median values for policy scenarios of 39% reduction in 2050, corresponding to annual energy efficiency improvement rates close to 1.6%. The IAM scenarios suggest median energy intensity reductions of passenger transport of 57–61% by the end of the century would align with warming levels of both 1.5°C and 2°C (C1–C4) given the corresponding decarbonisation of the fuels.

The scenarios in line with warming levels of 1.5°C or 2°C goals (C1 to C4) show different trends for freight’s energy intensity. The amount of overshoot and differences in demand for freight services and, to some extent, fuel carbon intensities contribute to these differences. For the two scenarios aligning with the warming levels of 1.5°C, the trajectories in 2030 and 2050 are quite different. The median C2 scenario that returns warming to 1.5°C (>50%) during the 21st century after high overshoot takes a trajectory with lower energy intensity improvements in the first half of the century. In contrast, the C1 scenario that limits warming to 1.5°C (>50%) during the 21st century with no or limited overshoot take on a more steadily declining trajectory across the means. The IAMs provide a less clear picture of required energy intensity improvements for freight than for passenger transport associated with different temperature targets. As for the carbon intensity of direct energy used across both passenger and freight, the modelling scenarios suggest very moderate reductions by 2030. The interquartile ranges for the C1 scenarios suggest global average reductions in carbon intensity of 5–10%. Across the other scenarios compatible with warming levels of 1.5°C or 2°C (C2–C4), the interquartile ranges span from 1–6% reductions in carbon intensity of direct energy used for transport. For 2050, the scenarios suggest that dependence on fuel decarbonisation increases with more stringent temperature targets. For the scenarios that limits warming to 1.5°C (>50%) during the 21st century with no or limited overshoot (C1), global carbon intensity of energy used for transport decreases by 37–60% (interquartile range) by 2050 with a mean of 50% reduction. The IAM scenarios in the AR6 database do not suggest full decarbonisation of transport fuels by 2100. The interquartile ranges across the C1–C4 set of scenarios, compatible with warming levels of 2°C and less, span from 61–91% reduction from 2020Mod levels.

Increasing the occupancy rate of passenger transport (Grubler et al. 2018) and reducing empty miles or increasing payload in freight deliveries (Gucwa and Schäfer 2013; McKinnon 2018) via improved logistics efficiency or streamlined supply chains (Mulholland et al. 2018), can present significant opportunities to effectively improve energy efficiency and decrease GHG emissions in transport. However, the recent trends of consumer behaviours have shown a declining occupancy rate of light-duty vehicles in industrialised countries (Schäfer and Yeh 2020), and the accelerating growing preference for SUVs challenges emissions reductions in the passenger car market (IEA 2019d). These trends motivate a strong focus on demand-side options.

Based on the scenario literature, a 51% reduction in median energy intensity of passenger transport and a corresponding 38–50% reduction in median carbon intensity by 2050 would be aligned with transition trajectories yielding warming levels below 1.5°C by the end of the century. For comparison, the LCA literature suggests a switch from current ICEs to current BEVs would yield a reduction in energy intensity well beyond 45% and up to 70%, for a mid-sized vehicle (Section 10.4). Correspondingly, a switch from diesel or gasoline to low-carbon electricity or low-carbon hydrogen would yield carbon intensity reduction beyond the median scenario value. Thus, the LCA literature suggests technologies exist today that would already match and exceed the median energy and carbon intensities values that might be needed by 2050 for low warming levels.

Two mechanisms for reducing carbon emissions from the transport sector are fuel switching for current vehicle technologies and transitioning to low-carbon vehicle technologies. Figure 10.21 combines data from IAMs and GTEMs on shares of transport final energy by fuel. These shares account for fuel uses across modes – road, aviation, rail, and shipping – and both passenger and freight transport. Since the technologies have different conversion efficiencies, these shares of final energy by fuel are necessarily different from the shares by service (passenger-km or tonne-km) by fuel and shares of vehicle stock by fuel. For example, a current battery electric LDV powertrain is roughly three times more energy-efficient than a comparable ICE powertrain (Section 10.3, Table 10.9 in Appendix 10.1); thus, fuel shares of 0.25 for electricity and 0.75 for oil could correspond to vehicle stock shares of 0.5 and 0.5, respectively. In general, while models may project that EVs constitute a greater share of road vehicle stock, and provide a greater share of road passenger-kilometres, their share of transport final energy (Figure 10.21) can still remain lower than the final energy share of fuels used in less-efficient (e.g., ICE) vehicles. Thus, the shares of transport final energy by fuel presented in Figure 10.21 should be interpreted with care.

IAM and GTEM scenarios indicate that fuel and technology shifts are crucial to reduce carbon emissions to achieve lower levels of warming (Edelenbosch et al. 2017; IEA 2017b). Across the transport sector, a technology shift towards advanced fuel vehicles is the dominant driver of decarbonisation in model projections. This trend is consistent across climate scenarios, with larger decreases in the final energy share of oil in scenarios that achieve progressively lower levels of warming. Due to efficiency improvements, the higher efficiency of advanced fuel vehicles, and slower progress in the freight sector, the final energy share of oil decreases more rapidly after 2030. By 2050, the final energy shares of electricity, biofuels, and alternative gaseous fuels increase, with shares from electricity generally about twice as high (median values from 10–30% across warming levels) as the shares from biofuels and gases (median values from 5–10%). While IAMs suggest that the final energy share of hydrogen will remain low in 2050, by 2100 the median projections include 5–10% hydrogen in transport final energy.

While few IAMs report final energy shares by transport mode or passenger/freight, several relevant studies provide insights into fuel share trends in passenger LDVs and freight vehicles. The IEA suggests that full LDV electrification would be the most promising low-carbon pathway to meet a 1.75°C goal (IEA 2017b). The MIT Economic Projection and Policy Analysis model focuses on the future deployment of gasoline versus EV technologies in the global LDV stock (Ghandi and Paltsev 2019). These authors estimate that the global stock of vehicles could increase from 1.1 billion vehicles in 2015 up to 1.8 billion by 2050, with a growth in EVs from about 1 million vehicles in 2015 up to 500 million in 2050. These changes are driven primarily by cost projections (mostly battery cost reductions). Similarly, the International Council on Clean Transport (ICCT) indicates that EV technology adoption in the light-duty sector can lead to considerable climate benefits. Their scenarios reach nearly 100% electrification of LDVs globally, leading to global GHG emissions from LDVs ranging from 0% to 50% of 2010 levels in 2050 (Lutsey 2015). Khalili et al.(2019) estimate transport stocks through 2050 under aggressive climate mitigation scenarios that nearly eliminate road transport emissions. They find the demand for passenger transport could triple through 2050, but emissions targets could be met through widespread adoption of BEVs (80% of LDVs) and, to a lesser extent, fuel cell and plug-in hybrid electric vehicles. Contrary to these estimates, the US Energy Information Administration finds small adoption of electrification for LDVs and instead identifies diffusion of natural gas-fuelled LDVs in OECD and, to a greater extent, non-OECD countries through 2040. This trend occurs in a reference and a ‘low liquids’ case, which lowers LDV ownership growth rates and increases preferences for alternative fuel vehicles. A comprehensive overview of regional technology adoption models across many methodological approaches can be found in Jochem et al. (2018).

In freight transport, studies indicate a shift toward alternative fuels would need to be supplemented by efficiency improvements. The IEA suggests efficiency improvements would be essential for decarbonisation of trucks, aviation, and shipping in the short-to-medium term. At the same time, the IEA suggests that fuel switching to advanced biofuels would be needed to decarbonise freight in the long term (IEA 2019d). Mulholland et al. (2018) investigated the impacts of decarbonising road freight in two scenarios: countries complying with COP21 pledges and a second more ambitious reduction scenario in line with limiting global temperature rise to 1.75°C. Despite the deployment of logistics improvements, high-efficiency technologies, and low-carbon fuels, activity growth leads to a 47% increase in energy demand for road freight while overall GHG emissions from freight increase by 55% (4.8 GtCO2- eq) in 2050 (relative to 2015) in the COP21 scenario. In the 1.75°C scenario, decarbonisation happens primarily through a switch to alternative fuels (hybrid electric and full battery electric trucks), which leads to a 60% reduction in GHG emissions from freight in 2050 relative to 2015. Khalili et al. (2019) also find substantial shifts to alternative fuels in HDVs under aggressive climate mitigation scenarios. Battery electric, hydrogen fuel cell, and plug-in hybrid electric vehicles constitute 50%, 30%, and 15% of heavy-duty vehicles respectively in 2050. They also find 90% of buses would be electrified by 2050.

This section provides an updated, detailed assessment of future transport scenarios from IAM, GTEMs, and NTEMs given a wide range of assumptions and under a set of policy targets and conditions. The scenario modelling tools are necessary to aggregate individual options and understand how they fit into mitigation pathways from a systems perspective. The scenarios suggest that 43% (30–63% for the interquartile ranges) reductions in CO2 emissions from transport (below modelled 2020 levels) by 2050 would be compatible with warming levels of 1.5°C (C1–C2 group). While the global scenarios suggest emissions reductions in energy supply sectors at large precede those in the demand sectors (Section 3.4.1), a subset of the scenarios also demonstrate that more stringent emission reductions in the transport sector are feasible. For example, the illustrative mitigation pathways IMP-REN and IMP-LD suggest emissions reductions of 80% and 90% respectively are feasible by 2050en route to warming levels of 1.5°C with low or no overshoot by the end of the century.

The scenarios from the different models project continued growth in demand for freight and passenger services, particularly in developing countries. The potential for demand reductions is evident, but the specifics of demand-reduction measures remain less explored by the scenario literature. This limitation notwithstanding, the IAM and GTEMs suggest that interventions that reduce the energy and fuel carbon intensity of transport are likely crucial to successful mitigation strategies.

The scenario literature suggests that serious attempts at carbon mitigation in the transport sector must examine the uptake of alternative fuels. The scenarios described in the IAMs and GTEMs literature decarbonise through a combination of fuels. Across the scenarios, electrification plays a key role, complemented by biofuels and hydrogen. In general terms, electrification tends to play the key role in passenger transport while biofuels and hydrogen are more prominent in the freight segment. The three illustrative mitigation pathways described in Box 10.4 exemplify different ways these technologies may be combined and still be compatible with warming levels of 1.5°C with low or no overshoot. Shifts towards alternative fuels must occur alongside shifts towards clean technologies in other sectors, as all alternative fuels have upstream impacts. Without considering other sectors, fuel shifts would not yield their full mitigation potentials. These collective efforts are particularly important for the electrification of transport, as the transformative mitigation potential is strongly dependent on the decarbonisation of the power sector. In this regard, the scenario literature is well aligned with the LCA literature reviewed in Section 10.4.

The models reviewed in this section would all generally be considered to have a good representation of fuels, technologies, and costs, but they often better represent land transport modes than shipping and aviation. While these models have their strengths in some areas, they have some limitations in other areas, like behavioural aspects. These models are also limited in their ability to account for unexpected technological innovation, such as a breakthrough in heavy vehicle fuels, artificial intelligence, autonomy and big data, even the extent of digital communications replacing travel (Section 10.2). As a result of these limitations, the models cannot yet provide an exhaustive set of options for decarbonising the transport sectors. These limitations notwithstanding, the models can find solutions encompassing the transport sector and its interactions with other sectors that are compatible with stringent emissions mitigation efforts. The solutions space of transportation technology trajectories is therefore wider than explored by the models, so there is still a need to better understand how all options in combination may support the transformative mitigation targets.

This final section draws some conclusions from the chapter and provides an overview-based feasibility assessment of the major transport mitigation options, as well as a description of emerging issues. The section ends by outlining an integrated framework for enabling the transformative changes that are emerging and required to meet the potential transformative scenarios from Section 10.7.

Transport is becoming a major focus for mitigation as its GHG emissions are large and growing faster than those of other sectors, especially in aviation and shipping. The scenarios literature suggests that without mitigation actions, transport emissions could grow by up to 65% by 2050. Alternatively, successful deployment of mitigation strategies could reduce sectoral emissions by 68%, which would be consistent with the goal of limiting temperature change to 1.5°C above pre-industrial levels. This chapter has reviewed the literature on all aspects of transport and has featured three special points of focus: (i) a survey of lifecycle analysis from the academic and industry community that uses these tools; (ii) surveying the modelling community for top-down and bottom-up approaches to identify decarbonisation pathways for the transport sector, and (iii) for the first time in the IPCC, separate sections on shipping and aviation. The analysis of the literature suggests three crucial components for the decarbonisation of the transport sector: demand and efficiency strategies, electromobility, and alternative fuels for shipping and aviation.

The challenge of decarbonisation requires a transition of the socio-technical system, which depends on the combination of technological innovation and societal change (Geels et al. 2017). A socio-technical system includes technology, regulation, user practices and markets, cultural meaning, infrastructure, maintenance networks, and supply networks (Geels 2005) (Cross Chapter Box 12 in Chapter 16). The multi-level perspective (MLP) is a framework that provides insights to assist policymakers when devising transformative transition policies (Rip and Kemp 1998; Geels 2002). Under the MLP framework, strategies are grouped into three different categories. The Micro level (niche) category includes strategies where innovation differs radically to that of the incumbent socio-technical system. The niche provides technological innovations a protected space during development and usually requires considerable R&D and demonstrations. In the Meso level (regime) state, demonstrations begin to emerge as options that can be adopted by leading groups who begin to overcome lock-in barriers from previous technological dependence. Finally, in the Macro level (landscape) stage, mainstreaming happens, and the socio-technical system enables innovations to break through. Figure 10.22 maps the MLP stages for the major mitigation strategies identified in this chapter.

Demand and behaviour. While technology options receive substantial attention in this chapter, there are many social and equity issues that cannot be neglected in any transformative change to mitigate climate change. Transport systems are socio-economic systems that include systemic factors that are developing into potentially transformative drivers of emissions from the sector. These systemic drivers include, for example, changes in urban form that minimise automobile dependence and reduce stranded assets; behaviour change programmes that emphasise shared values and economies; smart technologies that enable better and more equitable options for transit and active transport as well as integrated approaches to using autonomous vehicles; new ways of enabling electric charging systems to fit into electricity grids, creating synergistic benefits to grids, improving the value of electric transit, and reducing range anxiety for EV users; and new concepts for the future economy such as circular economy, dematerialisation, and shared economy that have the potential to affect the structure of the transport sector. The efficacy of demand reduction and efficiency opportunities depends on the degree of prioritisation and focus by government policy. Figure 10.22 suggests that innovative demand and efficiency strategies are at the regime scales. While these strategies are moving beyond R&D, they are not mainstreamed yet and have been shown to work much more effectively if combined with technology changes, as has been outlined in the transformative scenarios from Section 10.7 and in Chapter 5.

Electromobility in land-based transport . Since AR5, there has been a significant breakthrough in the opportunities to reduce transport GHG emissions in an economically efficient way due to electrification of land-based vehicle systems, which are now commercially available. EV technologies are particularly well established for light-duty passenger vehicles, including micromobility. Furthermore, there are positive developments to enable EV technologies for buses, light- and medium-duty trucks, and some rail applications (though advanced biofuels and hydrogen may also contribute to the decarbonisation of these vehicles in some contexts). In developing countries, where micromobility and public transit account for a large share of travel, EVs are ideal to support mitigation of emissions. Finally, demand for critical materials needed for batteries has become a focus of attention, as described in Box 10.6.

Electromobility options are moving from regime to landscape levels. This transition is evident in the trend of incumbent automobile manufacturers producing an increasing range of EVs in response to demand, policy, and regulatory signals. EVs for light-duty passenger travel are largely commercial and likely to become competitive with ICE vehicles in the early 2020s (Dia 2019; Bond et al. 2020; Koasidis et al. 2020). As these adopted technologies increase throughout cities and regions, governments and energy suppliers will have to deploy new infrastructure to support them, including reliable low-carbon grids and charging stations (Sierzchula et al. 2014). In addition, regulatory reviews will be necessary to ensure equitable transition and achievement of SDGs, addressing the multitude of possible barriers that may be present due to the incumbency of traditional automotive manufacturers and associated supporting elements of the socio-technical system (Newman 2020b) (Chapter 6). Similarly, new partnerships between government, industry, and communities will be needed to support the transition to electromobility. These partnerships could be particularly effective at supporting engagement and education programmes (Newman 2020b) (Chapter 8).

Deployment of electromobility is not limited to developed countries. The transportation sector in low- and middle-income countries includes millions of gas-powered motorcycles within cities across Africa, South-East Asia, and South America (Posada et al. 2011; Ehebrecht et al. 2018). Many of these motorcycles function as taxis. In Kampala, Uganda, estimates place the number of motorcycle taxis, known locally as boda bodas, at around 40,000 (Ehebrecht et al. 2018). The popularity of the motorcycle for personal and taxi use is due to many factors including lower upfront costs, lack of regulation, and mobility in highly congested urban contexts (Posada et al. 2011; UNECE 2018). While motorcycles are often seen as a more fuel-efficient alternative, emissions can be worse from two-wheelers than cars, particularly nitrogen oxides (NOx), carbon monoxide (CO), and hydrocarbon emissions (Vasic and Weilenmann 2006; Ehebrecht et al. 2018). These two-wheeler emissions contribute to dangerous levels of air pollution across many cities in low- and middle-income countries. In Kampala, for example, air pollution levels frequently exceed levels deemed safe for humans by the World Health Organization (Kampala Capital City Authority 2018; World Health Organization 2018; Airqo 2020). To mitigate local and environmental impacts, electric boda boda providers are emerging in many cities, including Zembo in Kampala and Ampersand in Kigali, Rwanda.

Bulawayo, the second-largest city in Zimbabwe, is also looking at opportunities for deploying electromobility solutions. The city is now growing again after a difficult recent history, and there is a new emphasis on achieving the Sustainable Development Goals (City of Bulawayo 2020a; City of Bulawayo 2020b). With these goals in mind, Bulawayo is seeking opportunities for investment that can enable leapfrogging in private, fossil fuel vehicle ownership. In particular, trackless trams, paired with solar energy, have emerged as a potential pathway forward (Kazunga 2019). Trackless trams are a new battery-based mid-tier transit system that could enable urban development around stations and that use solar energy for powering both transit and the surrounding buildings (Newman et al. 2019). The new trams are rail-like in their capacities and speed, providing a vastly better mobility system that is decarbonised and enables low transport costs (Ndlovu and Newman 2020). While this concept is only under consideration in Bulawayo, climate funding could enable the wider deployment of such projects in developing countries.

Fuels for aviation and shipping. Despite technology improvements for land-based transport, equivalent technologies for long distance aviation and shipping remain elusive. Alternative fuels for use in long-range aviation and shipping are restricted to the niche level. The aviation sector is increasingly looking towards synthetic fuels using low-carbon combined with CO2 from direct air capture, while shipping is moving towards ammonia produced using low-carbon hydrogen. Biofuels are also of interest for these segments. To move out of the niche level, there is a need to set deployment targets to support breakthroughs in these fuels. Similarly, there is a need for regulatory changes to remove barriers in new procurement systems that accommodate uncertainty and risks inherent in the early adoption of new technologies and infrastructure (Borén 2019; Sclar et al. 2019; Marinaro et al. 2020). R&D programmes and demonstration trials are the best focus for achievingfuels for such systems. Finally, there is a need for regulatory changes. Such regulatory changes need to be coordinated through ICAO and IMO as well as with national implementation tools related to the Paris Agreement (see Box 10.5). Long-term visions, including creative exercises for cities and regions, will be required, providing a protected space for the purpose of trialling new technologies (Borén 2019; Geels 2019).

Figure 10.23 sets out the feasibility of the core mitigation options using the six criteria created for the cross-sectoral analysis. This feasibility assessment outlines how the conclusions outlined in Section 10.8.1 fit into the broader criteria created for feasibility in the whole AR6 report and that emphasise the SDGs. Figure 10.23 highlights that there is high confidence that demand reductions and mode shift can be feasible as the basis of a GHG emissions mitigation strategy for the transport sector. However, demand-side interventions work best when integrated with technology changes. The technologies that can support such changes have a range of potential limitations as well as opportunities. EVs have a reliance on renewable resources (wind, solar, and hydro) for power generation, which could pose constraints on geophysical resources, land use, and water use. Furthermore, expanding the deployment of EVs requires a rapid deployment of new power generation capacity and charging infrastructure. The overall feasibility of electric vehicles for land transport is likely high and their adoption is accelerating. HFCVs for land transport would also have constraints related to geophysical resource needs, land use, and water use. These constraints are likely higher than for EVs, since producing hydrogen with electricity reduces the overall efficiency of meeting travel demand. Furthermore, the infrastructure needed to produce, transport, and deliver hydrogen is under-developed and would require significant R&D and a rapid scale-up. Thus, the feasibility of HFCV is likely lower than for EVs. Biofuels could be used in all segments of the transport sector, but there may be some concerns about their feasibility. Specifically, there are concerns about land use, water use, impacts on water quality and eutrophication, and biodiversity impacts. Advanced biofuels could mitigate some concerns and the feasibility of using these fuels likely varies by world region. The feasibility assessment for alternative fuels for shipping and aviation suggests that hydrogen-based fuels like ammonia and synthetic fuels have the lowest technology readiness of all mitigation options considered in this chapter. Reliance on electrolytic hydrogen for the production of these fuels poses concerns about land and water use. Using ammonia for shipping could pose risks for air quality and toxic discharges to the environment. The DAC/BECCS infrastructure that would be needed to produce synthetic fuel does not yet exist. Thus, the feasibility suggests that the technologies for producing and using these hydrogen-based fuels for transport are in their infancy.

Planning for integration with the power sector: Decarbonising the transport sector will require significant growth in low-carbon electricity to power EVs, and more so for producing energy-intensive fuels, such as hydrogen, ammonia and synthetic fuels. Higher electricity demand will necessitate greater expansion of the power sector and increase land use. The strategic use of energy-intensive fuels, focused on harder-to-decarbonise transport segments, can minimise the increase in electricity demand. Additionally, integrated planning of transport and power infrastructure could enable sectoral synergies and reduce the environmental, social, and economic impacts of decarbonising transport and energy. For example, smart charging of EVs could support more efficient grid operations. Hydrogen production, which is likely crucial for the decarbonisation of shipping and aviation, could also serve as storage for electricity produced during low-demand periods. Integrated planning of transport and power infrastructure would be particularly useful in developing countries where ‘greenfield’ development doesn’t suffer from constraints imposed by legacy systems.

Shipping and aviation governance: Strategies to deliver fuels in sufficient quantity for aviation and shipping to achieve transformative targets are growing in intensity and often feature the need to review international and national governance. Some authors in the literature have argued that the governance of the international transport systems could be included in the Paris Agreement process (Gençsü and Hino 2015; Lee 2018; Traut et al. 2018). Box 10.6 sets out these issues.

Managing critical minerals: Critical minerals are required to manufacture lithium-ion batteries (LIB) and other renewable power technologies. There has been growing awareness that critical minerals may face challenges related to resource availability, labour rights, and costs. Box 10.6 sets out the issues, showing how emerging national strategies on critical minerals, along with requirements from major vehicle manufacturers, are addressing the need for rapid development of new mines with a more balanced geography, less use of cobalt through continuing LIB innovations, and a focus on recycling batteries. The standardisation of battery modules and packaging within and across vehicle platforms, as well as increased focus on design for recyclability, are important. Given the high degree of potential recyclability of LIBs, a near closed-loop system in the future would be a feasible opportunity to minimise critical mineral issues.

Enabling creative foresight: Human culture has always had a creative instinct that enables the future to be better dealt with through imagination (Montgomery 2017). Science and engineering have often been preceded by artistic expressions; for example Jules Verne first dreamed of the hydrogen future in 1874 in his novel The Mysterious Island. Autonomous vehicles have regularly occupied the minds of science fiction authors and filmmakers (Braun 2019). Such narratives, scenario building, and foresighting are increasingly seen as a part of the climate change mitigation process (Lennon et al. 2015; Muiderman et al. 2020) and can ‘liberate oppressed imaginaries’ (Luque-Ayala 2018). Barber (2021) emphasised the important role of positive images about the future instead of dystopian visions and the impossibility of business-as-usual futures.

Transport visions can be a part of this cultural change as well as the more frequently presented visions of renewable energy (Wentland 2016; Breyer et al. 2017). There are some emerging technologies, like Maglev, Hyperloop, and drones that are likely to continue the electrification of transport even further (Daim 2021) and which are only recently at the imagination stage. Decarbonised visions for heavy vehicle systems appear to be a core need from the assessment of technologies in this chapter. Such visioning or foresighting requires deliberative processes and the literature contains a growing list of transport success stories based on such processes (Weymouth and Hartz-Karp 2015). Ultimately, reducing GHG emissions from the transport sector would benefit from creative visions that integrate a broad set of ideas about technologies, urban and infrastructure planning (including transport, electricity, and telecommunications infrastructure), and human behaviour and at the same time can create opportunities to achieve the SDGs.

Enabling transport climate emergency plans, local pledges and net zero strategies: National, regional and local governments are now producing transport plans with a climate emergency focus (Jaeger et al. 2015; Pollard 2019). Such plans are often grounded in the goals of the Paris Agreement, based around local low-carbon transport roadmaps that contain targets for and involve commitments or pledges from local stakeholders, such as workplaces, local community groups, and civil society organisations. Pledges often include phasing out fossil fuel-based cars, buses, and trucks (Plötz et al. 2020), strategies to meet the targets through infrastructure, urban regeneration and incentives, and detailed programmes to help citizens adopt change. These institution-led mechanisms could include bike-to-work campaigns, free transport passes, parking charges, or eliminating car benefits. Community-based solutions like solar sharing, community charging, and mobility as a service can generate new opportunities to facilitate low-carbon transport futures. Cities in India and China have established these transport roadmaps, which are also supported by the United Nations Centre for Regional Development’s Environmentally Sustainable Transport programme (Baeumler et al. 2012; Pathak and Shukla 2016; UNCRD 2020). There have been concerns raised that these pledges may be used to delay climate action in some cases (Lamb et al. 2020) but such pledges can be calculated at a personal level and applied through every level of activity from individual, household, neighbourhood, business, city, nation or groups of nations (Meyer and Newman 2020) and are increasingly being demonstrated through shared communities and local activism (Bloomberg and Pope 2017; Sharp 2018; Figueres and Rivett-Carnac 2020). Finally, the world’s major financing institutions are also engaging in decarbonisation efforts by requiring their recipients to commit to Net Zero Strategies before they can receive their funding (Robins 2018; Newman 2020a) (Chapter 15, Cross-Chapter Box 1 in Chapter 1). As a result, transparent methods are emerging for calculating what these financing requirements mean for transport by companies, cities, regions, and infrastructure projects (Chapters 8 and 15). The continued engagement of financial institutions may, like in other sectors, become a major factor in enabling transformative futures for transport as long as governance and communities continue to express the need for such change.

Using the right tools and strategies is crucial for the successful deployment of mitigation options. Table 10.7 summarises the tools and strategies required to enable electromobility, new fuels for aviation and shipping, and the more social aspects of demand efficiency.

Table 10.7 | Tools and strategies for enabling mitigation options to achieve transformative scenarios.

Electromobility is the biggest change in transport since AR5. When powered with low-carbon electricity, electric vehicles (EVs) provide a mechanism for major GHG emissions reductions from the largest sources in the transport sectors, including cars, motorbikes, autorickshaws, buses and trucks. The mitigation potential of EVs depends on the decarbonisation of the power system. EVs can be charged by home or business renewable power before or in parallel to the transition to grid-based low-carbon power.

Electromobility is happening rapidly in micromobility (e-autorickshaws, e-scooters, e-bikes) and in transit systems, especially buses. EV adoption is also accelerating for personal cars. EVs can be used in grid stabilisation through smart charging applications.

The state-of-the-art lithium-lon batteries (LIBs) available in 2020 are superior to alternative cell technologies in terms of battery life, energy density, specific energy, and cost. The expected further improvements in LIBs suggest these chemistries will remain superior to alternative battery technologies in the medium term, and therefore LIBs will continue to dominate the electric vehicle market.

Dependence on LIB metals will remain, which may be a concern from the perspective of resource availability and costs. However, the demand for such metals is much lower than the reserves available, with many new mines starting up in response to the new market, particularly in a diversity of places.

Recycling batteries will significantly reduce long-term resource requirements. The standardisation of battery modules and packaging within and across vehicle platforms, as well as increased focus on design for recyclability, are important. Many mobility manufacturers and governments are considering battery recycling issues to ensure the process is mainstreamed.

The most significant enabling condition in electromobility is to provide electric recharging opportunities and an integration strategy so that vehicles support the grid.

There are few obvious solutions to decarbonising heavy vehicles like international ships and planes. The main focus has been increased efficiency, which so far has not prevented these large vehicles from becoming the fastest-growing source of GHG globally. These vehicles likely need alternative fuels that can be fitted to the present propulsion systems. Emerging demonstrations suggest that ammonia, advanced biofuels, or synthetic fuels could become commercial.

Electric propulsion using hydrogen fuel cells or Li-ion batteries could work with short-haul aviation and shipping, but the large long-lived vessels and aircraft likely need alternative liquid fuels for most major long-distance functions.

Advanced biofuels, if sourced from resources with low GHG footprints, offer decarbonisation opportunities. As shown in Chapters 2, 6, and 12, there are multiple issues constraining traditional biofuels. Sustainable land management and feedstocks, as well as R&D efforts to improve lignocellulosic conversion routes, are key to maximising the mitigation potential from advanced biofuels.

Synthetic jet and marine fuels can be made using CO2 captured with DAC/BECCS and low-carbon hydrogen. These fuels may also have less contrails-based climate impacts and lower emissions of local air pollutants. However, these fuels still require significant R&D and demonstration.

The deployment of low-carbon aviation and shipping fuels that support decarbonisation of the transport sector will likely require changes to national and international governance structures.

In mid-2020, the IPCC, in collaboration with the Norwegian University of Science and Technology, released a request for data from the lifecycle assessment (LCA) community, to estimate the lifecycle greenhouse (GHG) emissions of various passenger and freight transport pathways. The data requested included information about vehicle and fuel types, vintages, vehicle efficiency, payload, emissions from vehicle and battery manufacturing, and fuel cycle emissions factors, among others.

Data submissions were received from approximately 20 research groups, referencing around 30 unique publications. These submissions were supplemented by an additional 20 studies from the literature. While much of this literature was focused on LDVs and trucks, relatively few studies referenced bus and rail pathways.

First, the datapoints were separated into categories based on the approximate classification (e.g., heavy-duty vs medium-duty trucks), powertrain (i.e., internal combustion engines (ICEV), hybrid electric vehicles (HEV), battery electric vehicles (BEV), fuel cell vehicles (FCV)), and fuel combination. For each category of vehicle/powertrain/fuel, a simplified LCA that harmonises values from across the reviewed studies was constructed, using the following basic equation:

Where:

•Lifecycle GHG intensity represents the normalised lifecycle GHG emissions associated with each transportation mode, measured in gCO2-eq per passenger-kilometre (pkm)or gCO2-eq per tonne-kilometre (tkm).

•FC is the fuel consumption of the vehicle in megajoules (MJ) or kilowatt hours (kWh) per km.

•P represents the payload (measured in tonnes of cargo) or number of passengers, at a specified utilisation capacity (e.g., 50% payload or 80% occupancy).

•EF is an emissions factor representing the lifecycle GHG intensity of the fuel used, measured in gCO2-eq MJ–1 or gCO2-eq kWh –1. A single representative EF value is selected for each fuel type. When a given fuel type can be generated in different ways with substantially different upstream emissions factors (e.g., hydrogen from methane steam reforming vs hydrogen from water electrolysis), these are treated as two different fuel categories. The fuel emissions factors that were used are presented in Table 10.8.

•VC are the vehicle cycle emissions of the vehicle, measured in gCO2-eq per vehicle. This may include vehicle manufacturing, maintenance and end of life, or just manufacturing.

•LVKT is the lifetime vehicle kilometres travelled.

Note: for plug-in hybrid electric vehicles (PHEV), the value of FC/P*EF is a weighted sum of this aggregate term for each of battery and diesel/gasoline operation.

Fuel emissions factors used are presented in Table 10.8. Note that the fuel emissions factors were compiled from several studies that used different global warming potential (GWP) values in their underlying assumptions, and therefore the numbers reported here may be slightly different if the 100-year global warming potential (GWP100) from the AR6 had been used. This difference would be small given the small contribution from non-CO2 gases to the total lifecycle emissions. For example, methane (CH4) emissions exist in the lifecycle of natural gas supply chains or natural gas-dependent supply chains such as hydrogen from steam methane reforming (SMR). Recent data from the US suggests emissions of approximately 0.2–0.3 gCH4 per MJ natural gas (Littlefield et al. 2017, 2019), which would range by no more than 1–2 gCO2-eq per MJ natural gas (<3% of natural gas lifecycle emissions) when converting from a GWP100 of 25 (AR4) or 36 (AR5) to the current (AR6) GWP100 of 29.8.

For LDVs, the entire distribution of estimated lifecycle emissions is presented for each vehicle/powertrain/fuel category (as a boxplot) in Figure 10.4. For trucks, rail and buses, only the low and high estimates are presented (as solid bars) in Figures 10.6 and 10.8, since the number of datapoints were not sufficient to present as a distribution. Table 10.9 presents the low and high estimates of fuel efficiency for each category. The references used are reported in the main text.

For transit and freight, the lifecycle harmonisation exercise allows two aggregate parameters to vary from the low to high among submitted values within each category: FC/P and VC/P. Aggregate parameters are used to capture internal correlations (e.g., fuel consumption and payload; both depend heavily on vehicle size) and are presented in Tables 10.10 to 10.14. The references used are reported in the main text.

Table 10.8 | Fuel emissions factors used to estimate lifecycle greenhouse gas (GHG) emissions of passenger and freight transport pathways.

Table 10.9 | Range of fuel efficiencies for light-duty vehicles by fuel and powertrain category, per vehicle kilometre.

Table 10.10 | Range of fuel efficiencies for buses by fuel and powertrain category, at 80% occupancy.

Table 10.11 | Range of fuel efficiencies for passenger rail by fuel and powertrain category, at 80% occupancy.

aOccupancy corresponds to average European occupancy rates (IEA 2019e).

Table 10.12 | Range of fuel efficiencies for heavy-duty truck by fuel and powertrain category, at 100% payload.

aAmmonia ICEV trucks are assumed to have the same fuel economy as diesel ICEVs due to lack of data.

bAmmonia FCV trucks are assumed to have the same fuel economy as hydrogen FCVs due to lack of data.

Table 10.13 | Range of fuel efficiencies for medium-duty truck by fuel and powertrain category, at 100% payload.

aAmmonia ICEV trucks are assumed to have the same fuel economy as diesel ICEVs due to lack of data.

bAmmonia FCV trucks are assumed to have the same fuel economy as Hydrogen FCVs due to lack of data.

Table 10.14 | Range of fuel efficiencies for freight rail by fuel and powertrain category, at an average payload.

For diesel, a range of USD0.5–2.5 per litre is used based on historic diesel costs across all OECD countries reported in the IEA Energy Prices and Taxes Statistics database (IEA 2021c) since 2010. The lower end of this range is consistent with the minimum projected value from the 2021 US Annual Energy Outlook (low oil price scenario, USD0.55 l –1) (US Energy Information Administration 2021). The upper end of the range encompasses both the maximum diesel price observed in the 2021 US Annual Energy Outlook projections (high oil price scenario, USD1.5 l –1) (US Energy Information Administration 2021), and the diesel price that would correspond to the 2020 IEA World Energy Outlook crude oil price projections (Stated Policies scenario) (IEA 2020b), assuming the historical price relationship between crude oil and diesel is maintained (USD1.5 l –1). For reference, the IEA reports current world-average automotive diesel costs to be around 1 USD l –1 (IEA 2021d). The selected range also captures the current range of production costs for values for bio-based and synthetic diesels (EUR51–144 MWh –1, corresponding to USD0.6–1.70 l –1), which are generally still higher than wholesale petroleum diesel costs (EUR30–50 MWh –1, corresponding to USD0.35–0.6 l –1), as reported by IEA (IEA 2020c). This range also encompasses costs for synthesised electrofuels from electrolytic hydrogen, as reported in Chapter 6 (USD1.6 l –1).

The range of electricity costs used here is consistent with the range of levelised cost of electricity estimates presented in Chapter 6 (USD20–200 MWh –1).

For hydrogen, a range of USD1 to USD13 per kilogram is used. The upper end of this range corresponds approximately to reported retail costs in the US (Eudy and Post 2018b; Argonne National Laboratory 2020; Burnham et al. 2021). Despite the high upper bound, lower costs (USD6–7 kg –1) are already consistent with recent cost estimates of hydrogen produced via electrolysis (Chapter 6) and current production cost estimates from IRENA (IRENA 2020). The lower end of the range (USD1 kg –1) corresponds to projected future price decreases for electrolytic hydrogen (BNEF 2020; Hydrogen Council 2020; IRENA 2020), and is consistent with projections from Chapter 6 for the low end of long-term future prices for fossil hydrogen with CCS.

The vehicle efficiencies used in developing the lifecycle cost estimates were derived from the harmonised ranges used to develop lifecycle GHG estimates and are presented in Tables 10.9 to 10.14.

For buses, a 40-foot North American transit bus with a passenger capacity of 50, lifetime of 15 years, and an annual distance travelled of 72,400 km based on data in the ANL AFLEET model (Argonne National Laboratory 2020) is assumed. Maintenance costs were assumed to be USD0.63 per km for ICEV buses and USD0.38 per km for BEV and ICEV buses, also based on data from the AFLEET model (Argonne National Laboratory 2020). For ICEV and BEV purchase costs, data from the National Renewable Energy Laboratory (Johnson et al. 2020) is used for bounding ranges (USD430,000 to 500,000 for ICEV and USD579,000 to 1,200,000 for BEV), which encompass the default values from AFLEET model (Argonne National Laboratory 2020). Note that wider ranges are available in the literature (e.g., as low as USD120,000 per bus in Burnham et al. (2021) and Harris et al. (2020)); but these are not included in the sensitivity analysis to avoid conflating disparate vehicles. For FCV buses, the upper bound of the purchase price range (USD1,200,000) represents current costs in the US (Argonne National Laboratory 2020; Eudy and Post 2020), and the lower bound represents the target future value from the US Department of Energy (Eudy and Post 2020).

For freight and passenger rail, powertrain and vehicle operation and maintenance costs in USD per km from the IEA Future of Rail report (IEA 2019e) (IEA Figure 2.14 for passenger rail and IEA Figure 2.15 for freight rail) are used as a proxy for non-fuel costs. The ranges span conservative and forward-looking cases. In addition, the range for BEV rail ranges encompass short- and long-distance trains – corresponding to 100–200 km for passenger rail, and 400–750 km for freight rail. Note that all values exclude the base vehicle costs, but they are expected not to be significant as they are amortised over the lifetime distance travelled. For freight rail, a network that is representative of North America is assumed, with a payload of 2800 tonnes per train (IEA Figure 1.17), assumed to be utilised at 100%, with a lifetime of 10 years, and an average distance travelled of 120,000 km yr –1. For BEV freight rail, the range in powertrain costs is driven by battery costs of USD250–600 kWh –1, while for FCV freight rail, the range in powertrain costs is driven by fuel cell stack costs of USD50–1000 kW–1. For passenger rail, a network that is representative of Europe is assumed, with an average occupancy of 180 passengers per train (IEA Figure 1.14), with a lifetime of 10 years, and an average distance travelled of 115,000 km per year.

Capital cost ranges vary widely in the literature depending on the exact truck model, size and other assumptions. For ICEVs in this analysis, the lower bound (USD90,000) corresponds to the 2020 estimate for China from Moultak et al. (2017), and the upper bound (USD250,000) corresponds to the 2030 projection for the US from the same study. These values encompass the full range reported by Argonne (Burnham et al. 2021). The lower bound BEV cost (USD120,000) is taken from 2030 projections for China (Moultak et al. 2017) and the upper bound (USD780,000) is taken from 2020 cost estimates in the US (class 8 sleeper cab tractor) (Burnham et al. 2021). The lower bound for FCV trucks (USD130,000) corresponds to the 2050 estimate for class 8 sleeper cab tractors from Argonne National Laboratory and the upper bound (USD290,000) corresponds to the 2020 estimate from the same study (Burnham et al. 2021). These values span the full range reported by Moultak et al. (2017) for the US, Europe and China from 2020–2030.

The analysis uses a truck lifetime of 10 years and annual distance travelled of 140,000 km based on Burnham et al. (2021). An effective payload of 17 tonnes (80% of maximum payload of 21 tonnes) is assumed based on reported average effective payload submitted by Argonne National Laboratory in response to the IPCC LCA data collection call. A discount rate of 3% is used, based on Burnham et al. (2021) and consistent with the social discount rate from Chapter 3. Maintenance costs are assumed to be USD0.15 km –1 for ICEV trucks and USD0.09 km –1 for BEV and FCV trucks, as reported in Burnham et al. (2021).