Battery Management Getting Competitive For EVs

The success or failure of future electric vehicles will depend on where and how those cars are used, as well as significant advances in battery materials, energy density, and some very complex battery management systems.

Battery power needs to be balanced, stored for extended times, and delivered to wherever it is needed most in real time. This is a huge challenge, and nearly everything in a vehicle today is being re-architected, from mixtures of various elements in the cathodes and anodes of the batteries to the layout, shape, and packaging of the batteries and battery modules. And all of this must be monitored and managed electronically to ensure the individual cells are aging consistently and charging properly.

Replacing internal combustion engines with battery-powered motors was just the first step in a complex technology shift. Increasing the range these vehicles can drive on a single charge, and reducing the time it takes to charge the batteries, is proving to be a much thornier challenge. Battery capacity has been increasing at a rate of about 5% to 6% a year. Denser storage and/or more batteries would allow vehicles to drive longer and accelerate faster, but the availability of materials needed to make these batteries is constrained by geopolitics and environmental concerns, so numerous alternatives are being considered.

The problem is how to make batteries hold more charge while also making them more resilient to rapid charging. “Today they’re blending a little bit of silicon inside the anode — something like 5% to 10% — and they’re looking at ways to blend in even more silicon,” said Felix Weidner, senior staff engineer at Infineon. “During charging, you pump these lithium ions into the graphene, but you need quite a lot of graphene for one lithium ion. The idea is that you will get more energy density with silicon because the silicon can catch more lithium ions. But it also comes with drawbacks. It’s not as stable, and it expands a lot. Otherwise you would use 100% silicon anodes.”

Two other elements, nickel and cobalt, are used today to increase density and reduce the risk of fire. But the specter of shortages caused by wars and limited sources has boosted the price of batteries and the electric vehicles that rely on them, and prompted a global search for new materials that are more plentiful and easier to work with.

“Cost is what limits people from buying electric vehicles today,” said Venkat Srinivasan, director of the Argonne Collaborative Center for Energy Storage Science and deputy director of the Joint Center for Energy Storage Research. “We’ve estimated that this year the cost of batteries was approximately $130 a kilowatt hour, and you need 90 kilowatt hours for a typical electric car. We think the target needs to be around $65 per kilowatt hour. But it looks like 2022 will be the first year where the average cost of batteries will go up — although not for every car company because some have long-term contracts — because of cobalt and nickel shortages.”

Fig. 1: Global Li-ion EV battery demand projection. Source: Argonne National Laboratory

Just adding more batteries is an expensive option typically reserved for the luxury car market, which is the only place cars with a 500-mile range are showing up. More or bigger batteries also have an advantage in charge time, because it’s faster to charge an empty battery module than a nearly full one. The bigger the battery modules, and the more of them, the faster it is to get another 200 or 300 miles of range because the most time-consuming part of that charging is when they are nearly full.

But range also can vary due to other factors. Wind drag, for instance, has become a key concern for wheel design. Ambient temperature also can impact range, and so can mountainous terrain.

A 2019 field test by Consumer Reports showed that range in electric cars was reduced by almost half in cold climates. The following year, Car and Driver cranked up the heat and heated seats in a Tesla Model 3 and reported that the car’s range dropped by 60 miles. A motor does not generate heat the way an internal combustion engine does, so it’s basically like attaching a small oven to the batteries. In hot weather, an air conditioner has a similar impact.

The more frequently batteries are charged, the shorter the lifespan of those batteries, too. This generally shows up in a reduction in maximum range over time, similar to what happens in a cell phone when the battery degrades over time. But replacing the batteries in a vehicle — and that needs to happen all at once in order to retain balance across a battery module and between modules — is much more expensive than buying a new phone. This, in turn, is forcing automakers to think much more like chipmakers, where the power budget is fixed and new electronics need to fit within that budget.

“The goal is to minimize the number of charging cycles, because the cost of replacing batteries is scaring some people away from the technology,” said David Fritz, senior director for autonomous and ADAS at Siemens Digital Industries Software. “The less power that we can have the system consume, the more efficiently we can drive the motors. We can use composite materials to help reduce the weight, but there’s only so much we can eliminate. The next big challenge is understanding what all the electronics do and how much they consume. That’s something we can control by turning these devices off and putting them in low-power mode. That has to happen before we make the leap from Level 3 to Levels 4 and 5.”

Battery management
From a purely functional standpoint, battery management is the next big differentiator for carmakers. While auto OEMs will continue to differentiate with flashy features, there is limited uniqueness and customization when it comes to motors and transmissions in electric vehicles.

Battery management is a lot harder than it sounds. Today, it involves heating, cooling, and determining the optimum percentage of charge to prolong the life of the batteries. But that’s just the beginning. In the future, battery architectures are expected to become much more sophisticated, possibly involving dedicated batteries for different tasks and new materials. There also are expected to be workarounds for bad cells, almost the way ECC memory does with DRAM, maintaining maximum range over longer periods of time.

A couple of metrics already exist for this. One is state of charge (SOC), which is how much energy is in a battery at any given moment. The second is state of health (SOH), which is what percentage of the battery capacity is available.

“State of health is going to change over time,” said Scott Winder, system application engineer at Infineon Technologies. “There are a few factors in play for that. One is that as you increase the charge near its maximum capacity, you get chemical changes within a cell. If you charge it fully, it may not last as long. That’s something cell phone manufacturers have been addressing. There are trip modes, where normally you charge the battery for 250-mile range, but for a long trip you may charge it to 300 miles. There are also heat issues due to resistance. It’s easier to charge a battery when it’s empty, but it’s more difficult when it’s almost full. So you can start faster, and then reduce the charge based on the temperature of the cells.”

Because batteries can overheat and cause fires, there also is a safety aspect to batteries. “The cell health is one of the most important indicators, because when you have a car fire it starts with a single cell — unless there is an accident that damages multiple cells,” Winder said. “Maybe it didn’t charge right or something happened and it heats to the point of thermal runaway. These are usually in metal boxes, which are there to try to control the spread, but there’s so much energy that eventually it’s going to spread to the rest of the vehicle.”

Companies working with electric vehicles are paying special attention to thermal issues inside the batteries, and this is one of the challenges with introducing new materials. Everything needs to be tested under extreme conditions over time, and in the case of automotive applications, that could be as long as five years.

“You need to monitor the batteries, and you need to decide what the status of the individual cells is,” said Roland Jancke, design methodology head at Fraunhofer IIS’ Engineering of Adaptive Systems Division. “There will be spare cells, but these systems need to decide when to switch over and what the health will be. So you need a management system and some kind of battery management chips. To create this you need full simulation, where you can inject faults into an overall battery pack and see what happens if one of these cells fails, and see what the battery management system does. Is it monitoring everything correctly? Are the diagnostics working? Is it switching over to a different cell?”

Battery architectures
Batteries are physically heavy, and they need to be packed into a vehicle in a way that improves handling of the vehicle. They also need to be arranged in a way that supplies sufficient energy to critical functions with minimal loss, while also avoiding overheating of the batteries.

“Right now, we have all of the batteries in a pack at the bottom of a vehicle, and various OEMs are using a lot of conductive cooling, just bringing the heat through the metal to the chassis,” said Infineon’s Winder. “There’s also liquid cooling being used, with conductive fluids. And if you spread the batteries out so that you have pieces integrated into the chassis at different points, you have a lot less concentration in one area. The other side of this is that we need to heat the batteries when the temperature outside is cold so they can operate at peak efficiency. By distributing the batteries, you need to heat them at various spots.”

But not all batteries are created equal, and not all of them are ideal for every task. Some systems in a vehicle are always on. Others may be used only occasionally. And response time for turning on the vehicle and looking in the backup camera need to be nearly instantaneous, while a delay of a second or two for starting up the infotainment system generally will go unnoticed. Whether that evolves into a distributed battery architecture, using different types of batteries, or better control of existing batteries, remains to be seen. But at this point, all options are being explored, including possibly using hydrogen fuel cells as a backup.

“Batteries are very good for light-duty vehicles like passenger cars,” said Argonne’s Srinivasan. “But when you go to heavy duty trucking, ships, and planes, there’s only so much energy density a battery can give you. People are starting to talk about fuels like hydrogen, which is carbon-neutral, or a carbon-free fuel like ammonia. When we start viewing de-carbonization in the context of different sectors, not just passenger cars, we will start to see other technologies play a role. But for passenger cars, cell phones, laptops, watches, whatever, there’s probably going to be some sort of battery.”

Fig. 2: Global reserves of key minerals used in batteries. Source: Argonne National Laboratory/USGS/U.S. Dept. of Energy, based on LiNi_0.8Mn_0.1Co_0.1O₂ cathode.

Designing for lower power
The other side of the power equation, of course, is to improve the efficiency of the chips and systems inside a vehicle.

“There’s going to be a bigger focus on smarter, more efficient electronics,” said Michal Siwinski, chief marketing officer at Arteris IP. “We’re already seeing some of that. Five years ago, automotive chips used to be done a certain way. Now, it’s all customized processes. Some of that is regulation and some of that is the reality that EVs are here to stay. But even as batteries get more advanced, you’re still going to have not hundreds, but thousands of different electronic subsystems and chips, and they’re all going to be connected. That will absolutely be a drain on the electrical supply.”

As with all complex electronics, one of the big challenges is figuring out how partition and prioritize power.

“The changes are comparable to the transition from multiple discrete ICs to a system-on-a-chip, said Siemens’ Fritz. “The way we get the power down in SoCs is with clock gating and turning stuff on and off when needed. In a car, that could be important, but from the perspective of car companies, that’s almost sacrilege. There are so many ECUs that are doing separate and individual tasks that they can’t possibly turn things on and off. We’re working with seven different OEMs, and each of them is taking very different approaches. One of them was looking at a Level 4 autonomy solution that’s going to take about 4 kilowatts. We were able to model that same solution, and based on state-of-the-art technology rather than some off-the-shelf, power-hungry x86 stuff, we were able to get it down to 40 watts. The total system, once you add all the peripherals, is 50 watts, versus 4 kilowatts. That has an impact on range and sustainability, because one full charge saves about 7 pounds of CO₂.”

Conclusion
Battery chemistry, battery management, and battery design are becoming increasingly complex. The holy grail remains a battery that can be charged quickly, last through hundreds of thousands of road miles, and which is both safe and relatively cheap. As with most complex electronics, this requires tradeoffs, some sophisticated architectures, and an almost constant rethinking of the power delivery and storage systems within a vehicle. So while features such as sound dampening and slick monitors are being used to entice buyers, there’s a lot of innovation and experimentation underway in places that most people never look.

“Carmakers have been trying to to separate the engine starting power from the infotainment and electronics, and have that on a separate electrical bus,” said Marc Swinnen, product marketing director for the semiconductor business unit at Ansys. “With electric vehicles, it’s different. These are big batteries — 60 to 70 kilowatt hours. For an electric vehicle, it’s all about range, and one of the big issues is heating. That may explain why people in Michigan and the Dakotas aren’t too keen on electric cars.”

And in the case of the whole electric car ecosystem, that remains a big challenge.