Skip to content
Join SEPA

The road to autonomous EVs: 7 questions utilities should start answering now

Someday — and possibly sooner than you think — your latest e-commerce purchase will be delivered to you by a drone, which will itself be launched from a convoy of self-driving electric delivery trucks.

Or maybe you’ll be commuting in constantly circulating shared autonomous electric vehicles (SAEVs), which stay on the road almost 24-7, and thus are very cheap to operate and can offer superlow fares.

Image: Waymo

The future of SAEVs and other autonomous electric transport is one of infinite, fascinating and complex possibilities — with many questions to answer and issues to explore. Bringing together a broad range of stakeholders to start the conversation was the impetus for a recent meeting in Washington, D.C., sponsored by the Smart Electric Power Alliance (SEPA) and the Association for Unmanned Vehicle Systems International (AUVSI).

Those at the event represented diverse perspectives on autonomous vehicle (AV) technology, how, where and how fast it will be adopted, and its ultimate impact on the grid. Stakeholders included autonomous vehicle research organizations, private consultants, automotive manufacturers, major package delivery companies, high-speed internet service providers, and electric utilities.

Even so, the level of consensus around the fast pace of adoption was dramatic and unexpected. The SEPA research team had anticipated at least some level of debate among attendees; walking out of the meeting, we realized that electric utilities should be in proactive planning mode — now.

Clearly, AVs face major uncertainties and legal and regulatory challenges. Some, such as the team at ReThinkX in a recent report, “Rethinking Transportation 2020-2030,” believe that widespread adoption of SAEVs could occur within 10 years after federal regulatory approval of autonomous vehicle technology – as soon as 2030. With such forecasts, it is absolutely critical electric utilities keep a pulse on these market trends.

While not nearly as dramatic, the obvious parallel here would be the growth curve of the residential and utility-scale solar markets. A decade ago, few could foresee the eventual expansion and mainstreaming of solar power, and the resulting impacts to the grid and utility business models. Similarly, electric AVs could, in the near future, become an integral part of daily life.

 

Why AVs will be EVs

The majority of attendees at the SEPA-AUVSI meeting predicted that autonomous vehicles (AVs) will be all-electric — even semi-trailer trucks. In the short term, some AVs may still be designed with traditional, internal combustion engines; however, electric AVs have numerous advantages. Electric vehicles (EVs) have a longer vehicle life expectancy — around 500,000 miles — while simultaneously requiring less maintenance. For example, they don’t need oil changes.

There is significant uncertainty about EV consumption forecasts — such as those from Bloomberg New Energy Finance — which currently project EV demand of 118 terawatt-hour (TWh) per year by 2030. However, ReThinkX suggests it could be as high as 733 TWh per year by 2030 due to SAEVs — roughly 18 percent of total forecasted electricity demand.

As mentioned in a recent SEPA report, “Utilities and Electric Vehicles: Evolving to unlock grid value,” utilities must lay the groundwork to adequately prepare for an electric vehicle future. This technology is not simple and more research is required to understand consumer behavior — particularly charging patterns both now and in the future. All aspects of utility planning should consider the impact of SAEVs in modeling scenarios.

Figure: Autonomous Electric Vehicle Fleet as a share of Electricity Demand

SAEV Adoption: Will it be urban or rural first?

Autonomous vehicles are expected to be adopted in urban areas first, before spreading to rural areas. The expansion of charging infrastructure and high consumer demand in cities will enable relatively quick integration of SAEVs. In contrast, rural areas — with low population density and long distances between passenger pick-ups — will make shared vehicles a harder sell.

Some at the SEPA-AUVSI meeting offered a counter scenario in which AV adoption starts in rural areas. Adapting and installing charging infrastructure in an urban setting can be costly and disruptive so may slow deployment. Whereas, in rural areas infrastructure could be built from the ground up. Continuing with the lower cost argument, journeys in rural areas are typically longer, so the miles driven, per trip and per passenger, would be higher. In this case, the economics of rural driving would actually be better than in urban areas.

However, what may ultimately tip the balance toward urban adoption is the simple fact that in densely populated, bustling cities, AVs could be in almost constant use. On the other hand, rural AVs could be travelling long distances with no passengers on board, which would reduce their cost effectiveness.

 

Will AVs be used for shared-ride fleets or personal use?

The attendees at the SEPA-AUVSI meeting agreed that the deployment of autonomous vehicles would fall into two main categories: fleets for shared rides and personal vehicles. While a combination of the two is likely, the exact mix is less clear. Specific uses for autonomous vehicles will change based on factors such as where you live or whether the car will be used for work commutes or family trips.

Based on the announcements from major AV manufacturers to date, shared rides will be the initial primary use for autonomous vehicles. Fleets of ride-sharing AVs could create inexpensive transportation, with vehicles in operation around the clock rather than sitting idly at home as is the case for any personal car. Fleet autonomous vehicles would be convenient to use and could be hailed or booked in a similar fashion to current ride-sharing services, such as Uber or Lyft. Trips could also be planned ahead in areas with lower population densities to minimize waiting and ensure that vehicles are not left empty between pickups.

Of course, in some circumstances, shared rides are neither feasible nor wanted. For example, more well-off consumers, who are less concerned about cost-per-mile economics, could be willing to spend more for a private, luxury vehicle, rather than use shared fleets.

RethinkX has predicted the market for private AVs would be relatively small — less than 5 percent of total AV sales. But as the price of these vehicles falls over time, more people could look to purchase their own. In addition, as the current, younger urban population grows older and moves from the city to suburbs, the demand for personal AVs could grow.

 

Will trucking go electric and autonomous?

Most conversations about autonomous vehicles overlook medium- and heavy-duty trucks, where a strong potential market exists. Specifically, the trucking industry is seeing a shortage of drivers, a situation that will likely become worse with time. With ground freight estimated to double by 2040, autonomous trucks could fill the gap between that burgeoning freight and ever-dwindling numbers of drivers.

One proposed solution is using convoys of autonomous trucks, with a human-driven lead vehicle at the front. These trucks could drive virtually around the clock, with a relatively small number of drivers necessary.

However, a major safety issue arises when autonomous trucks make left turns. This is due to attempting to cross oncoming traffic with their large size. AV manufacturing companies have been conducting tests in which autonomous delivery trucks are routed to only make right turns — which are easier and safer — and drones, taking off from larger trucks, deliver packages to customers’ homes.

Converting the American semi tractor-trailer fleet from diesel would also require a large amount of around-the-clock electricity. Fast-charging infrastructure will be critical — both off-highway stations and rest areas and centralized depot-style locations in more urban areas.

 

What are policy considerations?

In cities and states with supportive policies, the progression of AV adoption has been easier and faster. To speed up the process of autonomous adoption, barriers need to be removed at the national, state and city level. For example, at present, each state has different testing standards, and the rollout of charging infrastructure or AV incentives may depend on individual state or even city governments.

California is an example of a jurisdiction moving forward with approvals for AV testing. The California Office of Administrative Law has approved requests by 52 companies to conduct road tests of autonomous vehicles. The state has also passed new regulations that allow driverless testing and public use of autonomous test vehicles in February of this year.

To truly accelerate the adoption of AVs, the federal government will have to issue nationwide testing standards and other key policies. To date, no such action has occurred.

 

Will the electric load be centralized or distributed?

The anticipated rise of electric autonomous vehicles may require additional generation capacity, which in turn could heavily impact the grid. Shared ride fleets will likely rely on centralized charging locations, while personally owned vehicles will need a distributed charging network.

A medium-term solution to on-the-go personal AV charging could be implemented at a network of convenience stores, where charging stations could be operated by attendants who plug a cable into the waiting autonomous vehicles. Eventually, however, inductive chargers — which can recharge a battery wirelessly simply by driving over a metal plate — could become the norm.

 

How should we design rates to minimize grid impacts?

Introducing managed charging (also known as smart charging or V1G) can help minimize grid impacts by spreading out or aligning charge times with least-cost, clean electricity while also avoiding costly peaks. The current trend is to offer time-of-use rates aimed at incentivizing consumers to charge their vehicles during specific, off-peak time periods. However, in some jurisdictions, this approach has resulted in a second peak — also called a timer peak — when customers schedule their vehicles to be charged as soon as the off-peak rate begins.

Off-peak charging can potentially be more effectively managed with current chargers — that take several hours to charge a vehicle — using vehicle software for example. However, other types of charging, such as direct current (DC) fast chargers, will be far more difficult to control.

DC fast chargers today can recharge a car in approximately 20-30 minutes, but the next generation of these chargers could recharge in as little as 5-10 minutes. Management techniques for DC fast chargers will largely involve real-time price signals or technology solutions, such as demand response strategies or adding energy storage systems onsite at fast charging locations to minimize demand during periods of grid stress.

Regardless of which charging infrastructure ultimately prevails, utilities will need to play an active role in standards development, interoperability requirements, and rate design to fully leverage these autonomous EVs as a grid asset.

 

Communication, transparency and collaboration

While the SEPA-AUVSI meeting produced no specific set of recommendations or call to action, we all left with a better sense of the major change in transportation technology our country faces — potentially much sooner than any of us expect. For utilities, the first step could be laying the groundwork for an integrated network of charging stations that could be aggregated and managed, thus mitigating future costs to upgrade distribution and transmission systems.

The best way to ensure a future where SAEVs are a grid asset — and not a grid liability — can occur through greater communication, transparency and collaboration between stakeholders. SEPA will continue to facilitate these conversations through our Electric Vehicle Working Group.

For more information about our EV work, visit our Electric Vehicle Resources page.

Share