How physics-based models enable electric vehicles to last the distance and offer a better battery charging experience
EV sceptics continue to have doubts around battery life, range, and charging. The adoption of physics-based models within the battery management system can help to enhance all three and make electromobility more sustainable
Dr Christian Korte, Head of Software and Engineering, Breathe Battery Technologies
The internal combustion engine has been continually developed for more than a century and has reached a level of efficiency, performance, and robustness that the pioneering automotive engineers could scarcely dream of. All of this is thanks to step-changes in design, materials, and manufacturing – and of course, the introduction of key technologies such as high-pressure direct injection, turbocharging, and variable valve control.
Engines are also optimised to run on the wide range of fuels available worldwide, with their varying quality and purity levels, octane and cetane ratings, and biofuel content. As a result, whenever customers pull at the pump and refuel, they drive away knowing that their vehicle will run exactly the way they expect, each and every time, for as long as they own it.
By comparison, lithium-ion batteries are still in their infancy. Although developing at an astonishing rate – far greater than combustion engines ever have, illustrating the efforts expended to push them still further – the very best still have energy densities lower than that of gasoline or diesel fuels.
This is why it’s so important to make the most of the energy that can be stored in every cell, regardless of cell form factor or chemistry. Achieving this relies not only on the cells themselves and the battery pack as a whole, but also on the charging process. And that’s something that allows the electric vehicle (EV) driver – knowingly or otherwise – to exert far more influence on the battery’s behaviour than they ever could filling a combustion engine vehicle with liquid fuels.
It matters little to the combustion engine whether the ambient temperature is 10°F or 90°F as the hydrocarbons flow into the tank, and none at all on the level at the start or finish of the refuelling process. And if the vehicle manufacturer (OEM) recommends premium gasoline but the driver always goes for regular, transient response and full-load performance might be a little less, but it certainly shouldn’t accelerate engine wear. But fast charging can have a marked influence on battery life, and so can temperature and state-of-charge levels when the EV is plugged in.
OEMs and their battery suppliers know this and factor these variables into the design of the battery, its thermal management systems, and control software. This gives them confidence that the battery – mandated in the US with an 8-year/100,000-mile warranty, and 10-year/ 150,000 mile in California – will meet the needs of customers, not incur excessive warranty costs, and deliver a robust service life beyond that period.
But this is a challenge because most EV battery management systems still rely on look-up tables to control the charging process, just as engine ECUs do to control the combustion process – and have done for decades. Relying on maps of battery temperature, state-of-charge and state-of-health results in a stepped charging process that can deliver a considerable difference between how the battery actually performs and how it could perform if it was operated closer to the cells’ technical limits. The aim of operating well below the limits is to restrict the potential for harmful processes that reduce battery life or impact safety to occur. Lithium plating being one of the most important.
The high currents experienced during fast charging result in lithium building up on the anode’s surface. This can also occur when fast charging at low ambient temperatures; something OEMs try to prevent by pre-heating the battery, but this reduces energy efficiency. The more the lithium builds up, the more energy storage capacity falls – with a corresponding reduction in range over time, and, ultimately, battery life. The process also increases internal resistance, which not only increases charge times, but also reduces the maximum discharge rate, limiting the power available to the vehicle’s motors, and therefore performance.
Severe lithium plating results in the formation of lithium dendrites: if these become large enough, they can pierce the cell separator, leading to short circuits and potentially overheating within the battery or even thermal runaway. This will remain a challenge with solid-state batteries as well, as dendrites can also pierce solid electrolytes, which effectively perform the same function as separators.
The issue with stepped charging profiles based on look-up tables is that they’re not as intelligent, and can’t adapt as the cells age. Even the most complex look up tables, which lock the voltage to enable the current to adapt as the cell ages, only offer a pseudo-control on the overpotentials that guard against cell aging. This leads to a damaging positive feedback loop in which the control strategy, within the battery management system, fails to adequately account for the reduction in state-of-health over time, so charging becomes too aggressive, lowering state-of-health still further, and therefore widening the gap. This accelerates battery degradation and, ultimately, the point at which end-of-life is reached.
At Breathe, we’ve pioneered a completely different method. In place of look-up tables we’ve developed and patented a physics-based model which powers our adaptive charging software. It uses the same input parameters as the existing charging strategy, but runs in real-time and uses closed-loop control to estimate – with a high level of precision – the actual electrochemical states within the cells.. This active, intelligent approach mitigates harmful processes such as lithium plating and brings battery degradation under control because the charging process is always adapting to state-of-health, enhancing battery capacity retention and cycle life.
It’s also worth noting that traditional charging slows down significantly when batteries are partially charged, for example starting at 30% or 50%. However, physics-based models maintain more consistent charging speeds regardless of the starting state of charge, which is important given that customers in real-world use don’t always plug in when the SoC drops to 10% – the charge level typically quoted in product brochures. Also, in physics-based charging, the system performs reliably even when it’s cold or hot, unlike traditional charging strategies that are significantly impacted by temperature.
For consumers, this means that their batteries (which could be in everything from their EVs to their smartphones) benefit from a more consistent, reliable charging experience and less drop-off in range as the vehicle (or smartphone) ages. When it comes to EVs these attributes are essential in convincing sceptical consumers to make the switch away from the combustion engine vehicles they’re familiar with.
For OEMs, it means they can experience reduced warranty costs because batteries become more durable. They may also see enhanced residual values for their EVs as regulations, such as those being introduced in the EU from this year, require information of battery state-of-health and expected battery lifetime to become available to consumers.
Our partnership with Volvo Cars has demonstrated that we can reduce the time taken to charge from 10-80% by as much as 30% without affecting energy density or range. Performance enhancements are also not only limited to start of life, and we maintain faster charging speeds than existing charging protocols throughout the battery lifetime. With more OEMs quoting charging in miles of range gained in 15 minutes, this is an important metric – particularly as a good charging experience means more time driving and less time plugged-in.
Our competitors are looking to do the same, but the smart charging software they’re developing often remains focused primarily on state-of-charge and state-of-health estimations. These values are certainly important in gauging the battery’s condition, but their techniques don’t leverage the active modelling or control of factors that improve the charging experience: ours do, and that’s what OEMs want. And despite the inherent complexity of our model, we’ve designed it in such a way that it can be parameterised to suit an OEM’s chosen cell type in as little as six weeks. From there our software can be integrated directly into the battery management system, ready for testing and validation at the system level.
Given that batteries are by far the single most expensive element of any EV, it’s absolutely essential to extract the most value from them as possible before their capacity falls to the level where they’re ready to be repaired, begin second-life applications, or, finally, be recycled. This is essential if we’re to make the most sustainable use of the minerals within them and conserve natural resources. At the same time, EVs must also deliver the charging experience, long range, and durability and that makes it seamless for consumers to make the switch from combustion engine vehicles. We’re proud that our software is playing its part in making this possible.
READ the latest Batteries News shaping the battery market
How physics-based models enable electric vehicles to last the distance and offer a better battery charging experience
