vehicles. Only the Netherlands has attempted a roll-out of a nationwide road pricing policy, which was in 2010 (MuConsult, 2017). The draft plan included the option to differentiate according to environmental characteristics, and time and location of use (Geurs, 2010). The plan did not gain parliamentary approval and the new government in 2010 did not include road pricing in their government agreement (Government of the Netherlands, 2010). Since 2017, the Dutch government has the intention to run pilots for alternative vehicle taxation for passenger cars, such as road pricing, but it does not plan to implement a nationwide policy (Government of the Netherlands, 2017; Government of the Netherlands, 2018). The government does plan to implement road pricing for freight trucks, which is already in place in several European countries.

The United States has several state-level pilot projects to test road pricing. The test cases are clustered in the “I-95 Corridor Coalition”, with 13 states63 participating, and the Road Use Charges (RUC) West Coalition, which involves 14 states64 in the west. Delaware has been the main pilot state on the east coast with funding for two multi-year pilots, followed by a multi-state truck pilot in 2019 (I-95 Coalition, 2019a). Oregon is the most advanced, with policies in place that enable the implementation of road pricing, while Washington, California, Colorado and Minnesota have conducted pilots as well (OReGO, 2018; I-95 Coalition, 2019b).

Early consideration of the implications for tax revenue is important even if technological changes take time to percolate through the entire car fleet. An early start and a thorough preparation and discussion with stakeholders are important to institute reform at an appropriate extent and depth for the longer term challenges posed by transport decarbonisation (OECD, 2019c). Early preparation and gradual implementation helps reduce the risk of disruption and creates room for developing, designing and implementing the necessary accompanying measures and appropriate policy responses. In particular, due attention is required for the effects that tax reform could have for individuals and households of different incomes, and/or access to alternative transport modes other than individual vehicles.65

Shared and automated mobility

The emerging mobility revolutions of sharing and automation could significantly reshape road transport over the coming decades, with major implications for vehicle electrification. A transition to shared and/or autonomous vehicles and services with high utilisation rates could put EVs at a competitive advantage. Yet, several challenges remain for EVs to meet all the operational and technical requirements of such services.

63Maine, New Hampshire, Massachusetts, Pennsylvania, Vermont, New York, New Jersey, Delaware, Maryland, Virginia, North Carolina, South Caroline, Georgia and Florida.

64Oregon, California, Colorado, Hawaii, Washington, Utah, Arizona, Idaho, Montana, Nevada, New Mexico, North Dakota, Oklahoma and Texas.

65For example, accompanying policy measures may encourage the development of alternative travel modes (e.g. public transport) in the long-run or support those households that are affected disproportionally by the reform in the short-run, but cannot easily cope due to budget constraints.

PAGE | 196

IEA. All rights reserved.

Challenges

Despite the significant fuel and maintenance savings potential of EVs in shared fleets (car sharing, app-based ride-sharing and taxis), adoption of EVs in these fleets has been slow. For instance, EV shares on the major ride-sharing platforms remain below 1%, except Didi (1.3%) (Slowik et al., 2019). Several challenges must be overcome to spur electrification of shared mobility.

EVs are generally more expensive to purchase. While this could be overcome in well-managed fleets, it is a significant issue for vehicles owned by drivers, given that they are often capital constrained.

Another barrier to BEV adoption in these services is that few models available today meet all the specific operational requirements of taxis and ride-sharing services, which typically use vehicles in the medium and large market segments (where the availability of BEVs is currently limited) and include sufficient range (exceeding 250 km/day and possibly reaching 400 km, often including the use of auxiliary power loads such as air conditioning).

Taxis and ride-sharing fleets with range limitations and limited access to overnight/off-peak charging may face similar challenges, as searching for available chargers during busy periods could mean foregone revenues for drivers. Free-floating car sharing fleets also rely on public (and often fast) charging, requiring EVs with sufficient range, a well-designed charging network and/or user incentives to charge. Operational challenges for fleets can be further exacerbated by the combination of these issues.

The intensive use patterns of fleets also raises issues around battery durability. More frequent and more rapid charge cycles of fleet vehicles could degrade the battery more quickly (compared to private cars), adversely affecting range over the lifetime of the vehicle.

If (and when) fleets become highly automated, their even higher utilisation rates are likely to increase daily travel distances, requiring either larger and more expensive battery packs that are sufficient for an entire day of operations, or more frequent recharging (and downtime). Autonomous cars may also require significant power for on-board electronics. However, the efficiency of chips used in autonomous pilot vehicles is improving rapidly and has already dropped from 3-5 kW in the first generation to less than 1 kW (Dunietz, 2018; Hawkins, 2017; Gawron et al., 2018).

Since vehicle sharing and automation are innovations that will be adopted first and most widely in cities, and given concerns that ride-sharing services could lead to more vehicle travel and congestion, it is imperative that electric drive technologies are adopted rapidly in these new mobility services (Rodier and Michaels, 2019). Air pollution concerns in cities have prompted regulatory responses, with more and more cities restricting operations of vehicles that emit air pollutants. Realising rapid electrification will not only reduce local air pollution, but can promote EV adoption among private cars buyers and roll out of public charging infrastructure.

Solutions

Significant technology progress could help to address the challenges outlined. Urban driving offers optimal conditions for high fuel efficiency of EVs, thanks in part to the important contribution of regenerative braking to reduce fuel use per kilometre (IEA, 2019a). Higher rates of regenerative braking could also slow battery degradation rates (Keil and Jossen, 2015). Plugin hybrid electric vehicles may also help bridge the range issue in the near term. Battery

PAGE | 197

IEA. All rights reserved.

technology is improving rapidly, allowing newer EVs to achieve higher energy densities and longer ranges. Battery technology improvements may also alleviate durability issues. Recent case studies from Tesla show that this may already be less of a concern for certain battery designs (Lambert, 2019). Rapid cost declines in electric vehicles as well, realised through new manufacturing processes, may soon help EVs achieve parity, first on the basis of total ownership costs, and eventually even in purchase price, with conventional vehicles.

In the period where EVs reduce total ownership costs versus conventional ICE cars only after a few years of operation, high upfront costs may be less of a concern for fleets; mobility service providers that purchase and operate their own vehicles will both have greater access to capital than private car buyers, and high mileage will lead to faster pay back of initial capital costs. This is likely to strengthen the total cost of ownership advantage for EVs, providing them with a compelling business case at earlier stages and for a broader set of uses.

A combination of policy and company efforts could also accelerate the electrification of fleets. Uber’s Clean Air Programme in London works in concert with the city’s Ultra-Low Emissions Zone, providing financial incentives to drivers to switch to or drive more in EVs. To promote adoption of EVs among private owners who earn a living driving their vehicle as a taxi or for app-based ride-sharing companies (transport network companies [TNCs]), energy service companies (ESCOs) can play a role in offering financial solutions to clear initial cost barriers of EV adoption. Maven, a car sharing company, offers a service of short-term rentals of the Chevrolet Bolt BEV for USD 229 per week to drivers working for TNCs and other shared platforms. ESCOs could get into this game as well, follow similar EV leasing strategies with preferential terms for highly used mobility service cars. Utilities can design preferential contracts when installing home chargers or public chargers that can help spread costs of charging over time, leveraging the higher utilisation of mobility service cars.

Although the economics of electrifying ride-sharing vehicles are favourable, new policies and company efforts are key to accelerate the transition to all-electric vehicle fleets (Pavlenko et al., 2019). Ride-sharing fleet operators (e.g. Didi) are increasingly working with EV manufacturers to design purpose-built EVs in order to address issues around range, passenger capacity and cargo space.

Autonomous vehicle (AV) pilot projects are underway in over 80 cities around the world, and nearly all are using some form of electrified vehicle. In California, EVs now account for around 70% of automated vehicle trial miles, with PHEVs accounting for half of all miles. The reasons that AV pilot projects are overwhelmingly choosing EVs are numerous. For PHEVs there are benefits in cost, technology compatibility and reduced local pollutant emissions, which are likely to help AVs in their quest for wide public acceptance. The strong EV penetration and growing momentum to electric AVs suggests that progress in vehicle automation and electrification are likely to be self-reinforcing trends.

The intensive and distinct use patterns of shared and/or automated fleets implies higher (and different) needs for charging compared to private EVs. The availability and coverage of public and fast chargers could be a critical factor in how quickly these fleets become electric and how business models evolve around shared and/or automated mobility. In the near term, appropriate data sharing between policy makers, utilities and fleet operators could help anticipate needs for charging infrastructure as mobility service fleets electrify. Over the longterm, shifts towards shared autonomous electric vehicle fleets could improve the economics of charging infrastructure by increasing utilisation, promoting faster returns on investment and reducing reliance on subsidies and indirect revenue streams through grid services.

PAGE | 198

IEA. All rights reserved.