The Good And Bad Of Bi-Directional Charging

Challenges, possible solutions, and some intriguing economic models.


Auto OEMs are starting to offer bi-directional charging in EVs, allowing batteries to power homes during outages or wherever else it is needed, and to smooth out any hiccups in the grid. But this technology also can shorten the lifetime of batteries, and it can open the door to more cyberattacks.

The idea behind bi-directional charging is simple enough. EVs can store huge amounts of power, and that power can be as mobile as an automobile, or as localized as a car parked in a garage. Vehicle-to-grid charging has been under discussion for the better part of a decade, because when the grid gets overloaded, rather than instituting rolling blackouts, lithium-ion batteries can be used to make up the difference.

There is an added benefit in this, as well. The Natural Resources Defense Council estimates that 5% of electricity in the grid is lost through transmission and distribution, at a cost of about $6 billion per year. So reducing the distance that energy needs to travel makes it particularly attractive.

On average, automakers aim for 1,000 charges per battery system, according to industry sources. Using a rough calculation, at 300 miles per charge, that equates to a maximum of 300,000 miles. But the actual lifespan of batteries will vary, depending on whether automobiles are fully re-charged from empty, whether they are completely charged each time, or whether they are repeatedly topped off and at what level. How quickly they are charged also can have an impact, because fast charging is harder on batteries.

At least some of this can be improved through smart charging and battery management. “Because of artificial intelligence and machine learning algorithms, data analysis has become cheap,” said Venkat Srinivasan, director of the Argonne Collaborative Center for Energy Storage Science at Argonne National Laboratory. “So there is a sense that we can predict the cycle life of batteries better. The way we use to do this was to take a battery and charge and discharge it at different temperatures, and then you extrapolate and say, ‘Okay, at 55° my capacity is this, and it came down so much at 45°, so it’s going to last 15 years.’ That’s really a very rough cut, because if I change the way the battery is used, then all these rules of thumb go away. Now we have enough of these large neural nets that can allow us to interpolate. If you have enough data in the data set, where we look at different use cases and different ways the battery will charge and discharge at different temperatures, then we can figure out on the fly what the impact is of this bi-directional charging application. And in the future, we might be able to adjust the way we use the battery, the way we charge it, or the way we use it on the grid to minimize degradation over its life. That would be adaptively changing the way the battery gets used.”

Battery economics
Case in point: Pacific Gas & Electric, which serves northern and central California, has three pilot programs underway, in addition to collaborations with General Motors and Ford, to accelerate “vehicle-to-everything” technology. That includes providing backup power to homes and other buildings, as well as setting up community microgrids to support “temporary power,” or to store excess energy in those batteries when there is too much energy. The program includes incentives for program participants, as well, which needs to offset degradation in batteries connected to the grid.

Those incentives are meant to offset the cost of a reduced lifespan for batteries, but how quickly they will degrade due to cyclical aging isn’t entirely clear. There are ways to get more charge cycles out of a battery, but the price will go up.

“Car companies are trying to get to 1,000 cycles,” said Argonne’s Srinivasan. If you can do that, and get 300 miles every charge, then it can go 300,000 miles. That’s more than we typically use a vehicle. So to get to bi-directional charging, car companies may view this as a lifecycle. You may have two cycles a day, once because you’re driving around in the vehicle and the second cycle because you’re on the grid. That might be cost-effective for the consumer, because they’re using another good application to make money from it. And you can make batteries last more than 1,000 cycles if you want. It depends on how much energy density you can afford.”

Fig. 1: Bi-directional charging of grid. Source: Argonne National Laboratory

There is a sizable body of inconsistent data from researchers around the globe, including some data on ways to minimize damage to batteries, and different battery chemistries that can affect their lifespan. At present, replacement costs average about $20,000, according to multiple estimates, but that will likely change.

One reason for that change involves materials, some of which are more widely available and mainstream than others. Typically, the cathode (positive electrode through which electrons exit a cell) in EV batteries is made of nickel-cobalt-manganese, nickel-cobalt aluminum, or lithium iron phosphate. The first two of those materials are denser and can store more energy, while lithium iron phosphate is more robust. It’s also cheaper, partly because one of the main sources of nickel is Ukraine, where supplies have been disrupted by the ongoing war with Russia. Meanwhile, the anode (negative electrode through which electrons enter a cell) traditionally was made of graphite. But due to geopolitical rifts with China, battery makers and automotive OEMs have been adding more silicon into the anodes, which ultimately is expected to drive down the cost.

All of these materials need to be studied in real-world use cases under a variety of conditions to determine how many charges they can accept, and what the added cost will be to increase that number. And they all need to be weighed in light of new technology being developed to minimize more frequent charge/discharge cycles.

Different approaches
One such potential solution to battery degradation, according to Infineon, involves “a Totem-Pole PFC (TP-PFC) coupled with a Dual Active Bridge (DAB) topology, which allows the use of soft-switching and low device count while also offering high efficiency and the necessary galvanic isolation. These half/full bridges are connected to an inductor and high-frequency transformer that also sets the conversion ratio, respectively. The control method varies according to the demands of the application and the voltage range supported on either side. One typical approach is to control both bridges using a complementary pulse-width modulated (PWM) signal, typically from a microcontroller (MCU). Modification of the phase of the signals applied defines the direction of power transfer.”

Another common topology is the CLLC (two inductors and two capacitors) resonant converter. This approach is suited to high-frequency switching, thanks to its galvanic isolation and support of soft-switching. Because of this, smaller passives can be used. The use of capacitors on both sides of the full bridges supports bi-directional current flow.

Fig. 1: A power factor correction (PFC) with CLLC topology suited to bi-directional on-board charging supplied by a single-phase AC supply. Source: Infineon

Fig. 1: A power factor correction (PFC) with CLLC topology suited to bi-directional on-board charging supplied by a single-phase AC supply. Source: Infineon

According to an Infineon white paper, bi-directional power converters are well understood, but there are a range of challenges involved in bringing them to market in the context of the automotive space. “Reliability and functional safety are essential not only for automotive and electricity generation standards, but also for keeping the well-being of the driver and occupants. Finally, the result must also meet the demanding pricing common to all automotive applications.”

Security issues
With traditional charging, security concerns at the hardware level are generally limited to protecting sensitive data that lies within the vehicle’s operating system, such as passwords, usernames, and other credentials. With bi-directional charging, there is all that and more.

Questions about bidirectional charging security are likely to become more common as auto OEMs increasingly offer the technology in their vehicles and governments around the world develop charging infrastructure. Adam White, division president of Infineon’s Power and Sensor Systems division, said at a company event in October that bidirectional charging is an important trend in the automotive industry and is a particularly hot topic in Japan.

A handful of vehicles currently offer vehicle-to-grid (V2G), vehicle-to-home (V2H), or vehicle-to-load (V2L) capabilities. Earlier this year, the Department of Energy announced a memorandum of understanding aimed at accelerating vehicle-to-everything (V2X) technologies. “Integrating charging technology that powers vehicles and simultaneously pushes energy back into the electrical grid is a win-win for the future of clean transportation and our energy resilience overall,” said Deputy Secretary of Energy Dave Turk in a news release about the memorandum.

But experts say bidirectional technology could also cause harm from bad actors if certain security precautions aren’t taken, particularly with regards to the charger itself.

“When you get to bi-directional, most states require IEEE 1547, and that comes with a lot of advanced functionality that’s required including reactive power support,” said Jay Johnson, principal member of technical staff at Sandia National Laboratories. “Non-unity power factor that can provide reactive power is generally great because you can do voltage stability on the distribution system. But on the flip side, if that is compromised, you can manipulate the device and inject or absorb reactive power.”

In other words, says Johnson, a malicious party could potentially harness reactive power to damage the charger or deplete the vehicle’s battery, to name just a few possibilities. If the party compromised multiple bi-directional chargers attached to vehicles and injected power into the grid, that potentially could cause inter-area oscillations or other potentially damaging scenarios.

Luckily, vehicles have multiple layers of protection to defend against overcharging, so it is unlikely that a bidirectional charger could be used to create a vehicle battery explosion. But Johnson says there are personal safety concerns around the thermal elements of chargers. “For some of the very high-power charging devices, the cable is liquid cooled so that they can have smaller gauge wires to pump the same amount of current,” he said. “If the charger is compromised in a way that turns off the cooling system, the cable can get uncomfortably hot.”

This is all in addition to the existing security concerns that are present with single-direction charging. “The way chargers are constructed right now, they’re not doing a very good job at protection,” said Johnson. “Some electric vehicle intermediary devices will expose passwords, usernames, WiFi credentials or other things through website vulnerabilities, malicious firmware updates, or just extracting media directly and pulling off unencrypted data that way. Many of them have an SD card that can be plugged into another machine to try to figure out what kind of data is there. If you’ve got your WiFi password stored on that device it can connect back and give you updates on your phone. For another thing, if you’ve got credentials for doing SSH or VPN tunnels back into the cloud infrastructure, if it’s compromised an adversary could potentially communicate upstream and manipulate something going on in the cloud. That could then impact the entire fleet of devices connected to that company’s server. Those are the kind of scary things we’ve been seeing and it’s likely to continue for a few more years, at minimum.”

As for prevention, Johnson says charging manufacturers should consider disabling the physical ports on the devices that are often used for debugging, maintenance, or to allow users to check their charging status, because they can be easily compromised by hackers. Other important measures include implementing secure firmware updates and secure boot processes.

There has been much written about the advantages of localized energy generation and storage. As more homes take advantage of renewable energy, storing that energy in batteries at the local level is very attractive. There is less loss between the source of the energy being generated due to much shorter distances, and the energy stored is cleaner.

But there are tradeoffs that need to be considered, including the impact on batteries, the security of localized power connections, and the environmental impact of recycling the batteries when they no longer can hold a charge. There is much research in this area, but it will take time for solutions to emerge. In the meantime, there is plenty of activity, a lot of unanswered questions, along with some intriguing new possibilities that include different ownership models and batteries as a service.


Sandeep Dixit says:

Excellent article – good insights & ‘food for thought

Frank says:

What is missing is user acceptance. Who will sacrifice their battery charge for grid stability? It is meant for mobility.

Leave a Reply

(Note: This name will be displayed publicly)