EV Batteries Need Road Data, Not Laboratory Data
Electric vehicles depend on battery management systems that have been designed and tested under ideal laboratory conditions, and researchers are discovering significant discrepancies between lab and road inputs that impact algorithm performance.
As clean energy innovations find a place in existing markets, it becomes apparent that continued research and development will only further maximize the promise of these emerging technologies. The electric vehicle (EV) market is no different, and researchers are discovering that laboratory-trained software does not adequately account for differences among individual drivers or various road conditions.
On-the-road behavior affects battery performance in ways that can’t be tested in the lab. Image used courtesy of Adobe Stock
The operational role of software in the overall functioning of an EV is essential. EVs rely on battery management systems (BMS) that control diverse vehicle operations and report back to drivers on the health of the system.
The various functions of a battery management system. Image used courtesy of DMC
Core functionality depends on the BMS’s ability to predict and understand battery functionality and other critical equipment performance. The BMS is responsible for tracking battery life and giving drivers accurate information about how many more miles they can travel before recharging.
BMS performance is central to driver experience and also to the optimization of vehicle performance, but so far, BMS software is being designed in lab settings under ideal conditions that differ significantly from actual road experience. Recently, a team of Stanford engineers recognized a research opportunity to understand more about how individual driver differences and idiosyncratic road conditions impact battery performance.
The Importance of Real-World Data
Simona Onori and her colleagues at Stanford studied 3,750 hours of actual BMS driving data and battery performance, and the team concluded that there is a significant discrepancy between laboratory testing and battery performance on the road.
These differences are rooted in conditions and driver traits that might seem mundane, but they exert a considerable impact on battery performance. More research is needed to account for these differences and to improve BMS algorithms so that they more closely match real-world conditions rather than idealized lab environments.
Driver habits that exert the most impact on batteries include braking, acceleration, and deceleration, all of which inevitably alter the performance of a lithium-ion battery pack. Even one of these variables shows a statistical significance that cannot be accounted for in lab settings where batteries have been tested under ideal conditions. On the road, driving style and particularly hard acceleration drain the battery significantly, with as much as a 30% increase in energy consumption resulting from aggressive driving habits.
Local conditions related to climate also impact battery performance in surprising ways. The Stanford team tracked a correlation between electrical resistance and found it decreased in colder environments but increased in warmer conditions. So, it seems safe to conclude that battery health improves in a milder setting, suggesting that spring and summer battery performance will differ meaningfully compared to cold season performance.
Driver behaviors and temperature conditions are not the only important variables to consider, as drivers' charging habits also significantly affect the life of a battery pack. Frequent fast charging can degrade battery health and lead to a quicker decline.
The Cost Context
One of the main selling points for EVs is the cost savings for the driver, who will make fewer trips to the dreaded gas pump. However, without studying driver habits to improve the BMS and educating drivers on how these habits impact battery performance, there is a lost opportunity to maximize the savings used as a key selling point.
As seen in the figure below, the EV market is indeed growing.
Global EV sales figures. Image used courtesy of the International Council on Clean Transportation
Given this booming market and the potential to use EV performance to help reduce carbon emissions and protect drivers’ pockets, the Stanford team’s research seems even more important as a critical step toward maximizing the benefits of an already advantageous technology.
The story here is not just that we need more research to continue improving EV performance, but that field data is a necessary complement to lab numbers, which can accidentally exclude human idiosyncrasies like the lead foot some of us have when the light turns green.
A report on the Stanford team’s research is available on the university’s website.
