New Model Predicts Where Millions of Electric Cars Will Charge on Long Trips

Norway relies on 707 public fast charging stations to power thousands of long-distance electric vehicle journeys across rugged terrain and winding mountain passes. A team of researchers has developed a new computer model to predict future nationwide charging demand by linking real-world driver behavior with national transport and energy databases. The work maps out how millions of battery electric vehicles will strain power grids and roadside infrastructure over the coming years.

The global count of battery electric vehicles reached 58 million in 2024 and is projected to climb to roughly 390 million by 2035. Norway alone will need 9,000 public fast charging points by 2030, which is more than double the infrastructure available in 2021. However, rapid shifts in battery technology, grid limits, and driver habits make long-term infrastructure planning difficult. Most existing forecasting tools focus primarily on urban commutes and short-distance city driving, leaving a blind spot for rural highways and long inter-city trips where fast charging is essential.

To fill this gap, researchers Odd André Hjelkrem, Olav Kåre Malmin, Erlend Dahl, and Bendik Nybakk Torsæter built a modeling framework that simulates driver decisions along specific travel corridors. The study is published in Re-Generation in Transport, detailing how the team combined national traffic projections with survey data from the Norwegian EV Association, which surveyed approximately 10,000 electric vehicle owners about their travel habits and battery capacities.

How Drivers Are Simulated on Mountain Crossings

The model starts by estimating the state of charge of a vehicle battery, assuming drivers begin long journeys with an average battery fill level of 80 percent. Energy consumption is then calculated for every kilometer of the route, factoring in vehicle weight, powertrain efficiency, and terrain elevation. The authors factored in range anxiety by setting a threshold of 30 kilometers of remaining battery range, meaning drivers will actively seek a charger if their estimated reserve drops below that limit at the next stop or falls below 60 kilometers two stops away.

To demonstrate the method, the authors analyzed a 477-kilometer long-distance route between Bergen and Oslo. This corridor crosses mountain terrain with summits reaching 1,136 meters above sea level. The simulation tracked three distinct vehicle socket types, namely CHAdeMO, CCS, and Tesla, representing the fleet composition in Norway in 2020. Smaller battery packs of 20 kilowatt-hours required six separate charging stops along the Bergen-Oslo highway, whereas cars equipped with 70 kilowatt-hour packs only needed a single top-up before climbing the mountain pass. Regenerative braking recovered about six to eight percent of the total energy consumed during the journey.

Nationwide Projections and Future Grid Strain

Scaling the simulation to a nationwide level across all 707 public fast charging stations revealed striking changes between the baseline year 2020 and a projected 2030 scenario. Total daily energy demand across all stations is expected to jump from 68,357 kilowatt-hours in 2020 to 388,113 kilowatt-hours by 2030. The authors also tested two alternative 2030 scenarios where average battery ranges increase by 20 percent and 40 percent without adding battery weight. Those improvements reduced total energy demand to 299,192 kilowatt-hours and 225,685 kilowatt-hours per day, respectively, because fewer charging stops were required overall.

Validation against actual operational data collected from 13 fast charging stations over 23 months showed that the model captures real-world patterns reasonably well. The coefficient of determination, a statistical measure of how closely modeled data matches observed values, reached average values of 0.72 for energy demand and 0.76 for visiting traffic volume hourly profiles. Nevertheless, the authors noted that the model generally underestimated energy transfer and visiting vehicles during peak daytime hours, particularly around 14:00.

Limitations and Next Steps for Infrastructure Planning

The authors emphasized several limitations in their current approach that point toward future research avenues. The model currently relies on average annual daily traffic and does not account for day-to-day variations caused by holidays, weekends, or severe weather conditions. Ambient temperatures dropping well below freezing in winter or climbing during summer significantly increase energy demands for cabin heating and air conditioning, which alters charging behavior. Furthermore, the analysis omitted regional urban trips and potential queue delays at busy stations, which could force drivers to divert to alternative chargers during peak periods.

The research was published in Re-Generation in Transport on January 1, 2027.