Electric Cars Gain Traction, But Challenges Remain

Battery-powered electric vehicles are expected to reach a milestone in terms of shipments in 2019, but the technology faces several significant hurdles to gain wider adoption in the market.

Limited driving range, high costs, battery issues, and a spotty charging infrastructure are the main challenges for battery electric vehicles (BEVs). In addition, there are issues with various power semiconductors and other devices.

This helps explain why hybrid vehicles, which run on both battery and gasoline, today are more popular than battery-only electric cars. But carmakers and several China-based vendors are accelerating their efforts in the BEV market amid rapid growth in China and elsewhere. Worldwide production of battery-electric cars is expected to reach the 2 million mark in 2019, compared to 1.39 million vehicles produced in 2018, according to IHS Markit. By 2030, some 30% of all cars are expected to be electric, according to the International Energy Agency.

This sounds impressive, but electric vehicles represent only 1% to 3% of all passenger cars today. “It’s absolutely taking off with high compound annual growth rates, but it comes from a very small base,” said Guy Moxey, senior director of power products at Wolfspeed. “Still, the portion of battery-electric vehicles compared to the overall number of vehicles made is very small.”

Some regions are growing faster than others. For example, China, the world’s largest electric car market, has formulated a national policy around the technology due to environmental issues. China’s share of the electric vehicle market is expected to reach 57% in 2019, up from 55.5% in 2018, according to Frost & Sullivan. “If you walk out of the airport in Shenzhen, China, for example, 90% of all taxis are electric vehicles. By 2021, which is not far away, every single bus in China will need to be pure electric,” Moxey said. “China can steer the direction quite easily through government legislation. That’s not so easy to do in Europe or North America.”

Still, BEVs face several challenges to gain more traction in all regions. “While they may be great in terms of sustainability and having a lower impact on the environment, there are some trade-offs involved,” said Jim Hines, an analyst with TechInsights. “For most people, when they think of battery-electric vehicles, they are thinking short ranges, high costs, and maybe I need to have a charging station installed in my home. There are pre-conceptions that probably prevent certain consumers from even considering an electric vehicle.”

Some of those perceptions are off-base, while others are not. Nonetheless, BEV makers face many of the same technical and cost challenges as traditional vehicles. “There are some high hurdles to overcome in terms of reliability, qualification and functional safety. And there are some relatively high cost pressures,” Hines said.

All told, BEVs and the infrastructure must improve. Otherwise, it will remain a niche market. Even market leader Tesla faces some headwinds amid cost, quality and profitability issues.

Figure 1: Growth of car market. Source: IHS

Battery, charging issues
BEVs have been around for decades. The first modern electric car appeared in 1996, when GM launched the EV1, which was dropped in 2002. In 1997, Toyota introduced the Prius, a hybrid.

Over time, electric cars have evolved, and today consumers have several choices. BMW, Daimler, Ford, GM, Nissan, Tesla, Toyota, VW and others are investing billions of dollars in electric vehicles and are expanding their efforts in the market. In one example, Telsa recently introduced the Model Y, an SUV that sells for $39,000 for a vehicle with a range of 230 miles, and $47,000 for a range of 300 miles.

BEVs are different than traditional cars, which are powered by an internal combustion engine (ICE). Powered by lithium-ion batteries, BEVs consist of three main power blocks—an on-board charger, a DC-to-DC converter, and a traction inverter.

Figure 2: Inside a battery-electric car. Source: Rohm

The traction inverter converts energy from the batteries to the traction motor, which propels the vehicle. The on-board charger replenishes batteries from a power grid. A DC-to-DC converter steps down the power to lower voltages. This is used to control the doors, heater and windows.

An electric motor is more efficient than an ICE, but that isn’t a big selling point for BEVs. “One key factor is consumer acceptance,” said Ian Fletcher, an analyst at IHS Markit. “This tends to be fed into by other factors, though, such as the entry cost of the technology without subsidies or some degree of preferential treatment of EVs to make them attractive. The limitations presented by the technology also remain a concern for many customers, with not only range but the time taken to charge an EV when required as well as a still sparse charging infrastructure in some markets.”

Meanwhile, as before, the top technical challenge is the lithium-ion battery in BEVs. Today’s lithium-ion technology is reaching its limit. Other battery technologies are in R&D and not expected for some time.

“The energy density of the lithium-ion battery has nearly quadrupled in its 28 years of existence on the market through evolutionary improvements in materials and design,” said Philippe Vereecken, a member of the technical staff at Imec. “The energy density of Li-ion batteries currently can provide a limited driving range of 400 to 500 km (249 to 311 miles), whereas the consumer wants a driving range of 700 km (435 miles) or more. Also, the high cost of Li-ion batteries makes the EV expensive.”

In ICE-powered cars, it’s measured by miles per gallon. The energy consumption of a BEV is measured in kilowatt-hours per 100 miles (kWh/100 miles). “On average, it takes 30 kilowatt hours of electricity to go 100 miles. A kilowatt hour is the amount of electricity that is used if you are delivering 1,000 watts for one hour,” said Mitch Van Ochten, an applications engineer at Rohm.

BEVs consist of a battery pack with lithium-ion cells. To extend the driving ranges, OEMs are increasing the energy levels of the battery packs. For example, Tesla’s Model S has a 100 kilowatt-hour (kWh) battery pack, compared to 60 to 70kWh at one time.

“The capacity of the batteries is measured in kilowatt hours (kWh), which is a metric of energy. The more the energy density can be increased in batteries, the higher the energy capacity for a given battery,” TechInsights’ Hines said. “For an EV, the design of a battery pack is limited by the size and mass of the pack. There is a certain amount where it’s not feasible to make it any bigger, because we are adding so much mass to the vehicle. The other performance metrics begin to suffer. Mass is the enemy of handling, acceleration and braking. The greater the mass, the harder it is to achieve good results on those performance metrics.”

There are other issues. “The elephant in the room, however, has always been the cost of these complex battery packs, which had caused EV sticker prices to be meaningfully higher than traditional ICE-powered counterparts,” said Jed Dorsheimer, an analyst at Canaccord Genuity.

Battery cell costs have fallen from $1,000/kWh in 2010 to $200/kWh in 2018, analysts said. “Our expectation is that when Li-ion pricing is widely available near the $100/kWh pack threshold, the EV penetration rate will exponentially inflect,” Dorsheimer said.

That solves one problem. The next big challenge is the charging infrastructure. For EVs, there are three levels of charging—Level 1, Level 2 and Level 3. In Level 1, the car is charged by plugging the vehicle into a 120-volt AC home outlet via an on-board charger. On average, this takes 17 hours to charge a car, according to Canaccord Genuity.

In Level 2, the vehicle is plugged into a 240-volt power source at home or an outside charging station. This takes 3.5 to 7 hours, according to the firm.

Level 3 involves a standalone DC fast charging unit based on a 480-volt system. Charge times are faster, but these charging units are not geared for home installation. Instead, consumers must take the vehicle to a standalone charging station, much like taking a car to a gas station.

In the United States there currently are not enough fast charging stations. In total, there are 150,000 gas stations in the U.S., compared to 10,000 public EV charging stations, according to Canaccord Genuity.

This presents a problem for consumers, with some psychological implications, as well. “The infrastructure is quite important,” said Llewellyn Vaughan-Edmunds, director of strategic marketing at Applied Materials. “Right now, there is an anxiety called an ‘anxiety mileage range.’ There is a paranoia that the battery is going to run out.”

This is less problematic around town in some cities. “I’d say 60% of the population with electric vehicles travel within 15 miles every day. You can get to work and back every day with a charge with no problem,” Vaughan-Edmunds said.

The big issue is long-distance travel, where charging stations are not always available. In response, Tesla, private companies and consortiums are installing fast charging stations throughout the U.S. This, of course, takes massive investments.

“Everyone is focused on fast charging to try and get that anxiety of long-distance travel down,” Vaughan-Edmunds said. “They are trying to enable electric vehicle adoption rates more quickly by setting up more support charging stations. But instead of relying on a public utility or the government board in setting up electric vehicle chargers, they are setting up their own initiatives as a group or a consortium. The idea is to collaborate, make the electric vehicle chargers, and place them in the right locations. What this does is encourages the public to start buying electric vehicles.”

Reliability and efficiency
Batteries and charging issues aren’t the only challenges. Improving the efficiency, reliability and the cost of the sub-systems and devices are also critical.

Assuring reliability in automotive electronics is critical for all cars, particularly BEVs. “How we design the car, how we interact with the car, and how the car interacts with other devices affects the underlying semiconductors under the hood and in-cabin,” said Steven Liu, senior vice president of marketing at UMC. “Many new electrical parts are being upgraded and added into major sub-systems, such as ADAS, infotainment and the electrical power train.”

There are other reliability issues. “One of the biggest challenges for EVs and hybrids is how the microcontroller can optimize the power efficiency for all of the different components inside the EV, from high- to low-end designs to ensure long-term design flexibility,” Liu said. “Power conversion systems are essential and important to modern EVs. Robustness and reliability of the integrated power devices are key challenges for automotive power IC designs and manufacturing. Also, on-chip memory solutions need to comply with the AEC-Q100 standard in order to satisfy the stringent operating temperature specifications.”

Then, some OEMs are incorporating chips at advanced nodes, which has other implications. “When you have years to debug your process, you’re naturally going to have higher reliability,” said Jay Rathert, senior director of strategic collaborations at KLA. “But when you’re putting 7nm and 10nm parts in there, those processes still have a lot of maturing to do. There are still a lot of systematic defects and integration challenges that haven’t been debugged yet.”

Making the car more efficient is also critical. For this, the industry is focusing on the three main power blocks in a system—the on-board charger, the DC-to-DC converter, and the traction inverter.

OEMs use power semiconductors and other components in the power blocks. Power semiconductors are designed to boost the efficiencies and minimize the energy losses in systems. They are specialized transistors that operate as a switch, allowing the power to flow in the “on” state and stop it in the “off” state.

The dominant power semiconductor types are based on silicon, namely the power MOSFET and the insulated-gate bipolar transistor (IGBT). Power MOSFETs are used in applications up to 900 volts. IGBTs, the leading midrange power semiconductor, are used for 400-volt to 10-kilovolt applications.

Both types have some limitations. “When you go from 600 to 900 volts, silicon MOSFETs are good but they start losing some steam. IGBTs are good heavy lifters, but they are not quick or efficient,” Wolfspeed’s Moxey said.

That’s why the industry is interested in two wide band-gap technologies—silicon carbide (SiC) and gallium-nitride (GaN). Compared to silicon-based devices, GaN- and SiC-base power chips are faster and help eliminate the power losses in systems. GaN- and SiC-based devices are more expensive, however.

“An IGBT is a switch. A silicon carbide MOSFET is a switch. A silicon carbide MOSFET can switch a lot faster than an IGBT. That transition between off and on is a lot faster, so you waste a lot less power. So you get a highly efficient switch. Silicon carbide is a wide band-gap material. That’s the difference between wide band-gap and silicon. It’s the switching performance,” Applied’s Vaughan-Edmunds said.

In BEVs, the goal is to boost the efficiencies in the power blocks. “All of the blocks consume electricity. They convert electricity, but they also waste electricity. They are not 100% efficient. When you have a battery with a fixed amount of power, you don’t want to waste anything,” Wolfspeed’s Moxey said. “If you implement silicon carbide in the DC-to-DC converter or the on-board charger, for example, they are 1% to 2% more efficient than silicon. Not only are they more efficient, but also for the same power rating, the power density is higher. They are smaller. On the car, size and weight are a huge thing.”

GaN- and SiC-based power devices are making inroads in BEVs, but they are not widely used in some power blocks. For example, the traction inverter incorporates a network of power semiconductors. OEMs mainly use IGBTs, which are cost-effective solutions here.

The exception is Tesla’s Model 3, which is using SiC MOSFETs for the traction inverter. Other OEMs are exploring SIC MOSFETs here, although most are not jumping on the bandwagon due to cost considerations. “Currently, IGBTs are used a lot in traction inverters. Silicon carbide is like the premium solution. What’s going to happen with silicon carbide is as cost comes down, you will see it in the mainstream,” Applied’s Vaughan-Edmunds said.

It’s a different story for the on-board charger. This unit has three sub-systems—an input stage, power factor correction (PFC) circuitry, and a DC-to-DC converter. (This converter is different than the DC-to-DC converter power block.)

The PFC circuit shapes the input current and then maximizes the power level in the system. The PFC incorporates power semiconductors, diodes and other components. “You take an AC source from the wall and obtain a controlled DC voltage. You have to take the AC and power factor correct it into DC. Then, you use a DC-to-DC converter to regulate the output DC voltage,” Wolfspeed’s Moxey explained.

For the switch, OEMs use super-junction power MOSFETs. These devices, which are souped-up power MOSFETs, are suitable for these tasks. However, OEMs incorporate SiC diodes, not silicon diodes, in the PFC. SiC diodes don’t speed up the charging times, but they make the charger more efficient and reduce the size of the components.

“A diode is a device that passes electricity in one direction and blocks it in the opposite direction. There is at least one or sometimes two diodes involved in that power factor correction circuit. Your choice of diode influences the efficiency of that power factor correction circuit,” Rohm’s Van Ochten said. “There is no such thing as perfect efficiency from input to output. The same thing is true for power factor correction. If you want the best efficiency, you use silicon carbide diodes in that power factor correction circuit. A least 95% of people building PFC circuits choose silicon carbide diodes in that circuit. They alone will get you a couple of tenths of percent better efficiency than traditional silicon diodes.”

From there, the DC-to-DC convertor takes the DC voltage created by the PFC. “The second half of the charger takes the DC voltage that was created by the PFC and it transforms into the appropriate voltage for the battery you are working with. In addition to transforming it to the appropriate voltage, it also isolates the battery and its circuitry from the mains circuit,” Van Ochten said. “This transformer isolates the vehicle from the mains for safety purposes and it changes the voltage to the appropriate level, whether you have a 400-volt battery or an 800-volt battery. It does this through a transformer.”

This is where GaN fits in, at least for some OEMs. “The DC coming out the power factor correction is going to have to be switched on and off quickly to drive the transformer. It must be AC to go through the transformer. That AC may be 50-kilohertz or 200-kilohertz or maybe even a megahertz,” he said. “You can’t drive silicon carbide or super-junction MOSFETs at a megahertz, but you can drive GaN at that frequency. That’s what steers people to pick GaN so they can use a higher frequency AC into that transformer and use a smaller transformer. So, GaN would drive the primary of the transformer. The faster you can switch that DC on and off, the smaller you can make that transformer. You can push to a megahertz and maybe beyond.”

Self-driving cars?
The traction inverter and charger are a work in progress in BEVs. Then, for all cars, the next big things are self-driving cars and advanced driver-assistance system (ADAS) technologies.

Self-driving cars are still in R&D, but ADAS is already here. ADAS involves various safety features in a car, such as automatic emergency braking, lane detection and rear object warning.

In the ADAS world, “Level 1” involves the automation of one or more control functions in a car, while “Level 2” is the automation of two or more functions. Level 3 and 4 involve more self-driving capabilities. “Level 5” is fully autonomous, steering-wheel optional.

The more advanced cars are Level 2. This includes Tesla’s ADAS technology, dubbed Navigate on Autopilot. This allows a driver to navigate freeway interchanges on its own and doesn’t require lane change confirmation, but the driver needs to hold or supervise the wheel.

“It’s still something I would characterize as a next-generation driver assistance feature,” TechInsights’ Hines said. “It still requires the driver to be involved in a supervisory capacity. I’ve experienced it myself. You have to periodically touch the wheel to let it know you are there. But it basically does everything quite well, especially on the freeway. It will also do off-freeway. On the freeway and stop-and-go traffic, that’s where Tesla owners love it.”

Self-driving features may appeal to some consumers. Connectivity is a selling point for BEVs. “WiFi plays a major role in the infrastructure of in-car connectivity, such as HD video streaming, camera, display sharing and software updates,” UMC’s Liu said. “Bluetooth delivers high-fidelity voice and streaming audio support.”

Eventually, 5G will enter the picture with sub-6GHz capabilities, followed by mmWave. “For the out-car connectivity, such as 5G will need auto grade 1 at 28GHz, and 38GHz for high speed and high bandwidth communications. In the mmWave front, as CMOS geometries become smaller, Fmax becomes higher and able to serve higher frequency with better costs,” Liu added.

But for BEVs, consumers are concerned about sticker prices, not to mention battery ranges and charging issues. Even that might not be enough. What will help propel demand? “There are a couple of factors here,” TechInsights’ Hines said. “One is consumer education and market evangelism around electric vehicles that needs to take place.”

Still, BEVs will continue to make inroads, although perhaps not as fast as many have projected. But clearly, EVs are no longer a novelty item. They are here to stay.

Related Stories

China’s Electric Car Ambitions

SiC Chip Demand Surges