Electric vehicles

02Apr 2016

For a seemingly simple device with only two electrical connections to it, a battery is deceivingly misunderstood by the broad population, especially as batteries are now a common fixture in our technology-laden daily lives. I will highlight in this post five common misconceptions about the lithium-ion battery:


Well, not literally, but the acrobatic move captures the perceived hypersensitivity of the average consumer about past or secret special recipes that can help your battery. One of the silliest one I ever heard was to store the battery in the freezer to extend its life. PLEASE, DO NOT EVER DO THIS! Another silly is to charge the battery once it drops below 50%, or 40% or 30%….Let me be clear, you can use your phone down to zero and recharge it, and it will be just fine.

It is also now common to find apps that will “optimize” your battery. The reality is they do nada! Don’t bother.  Don’t also bother with task managers; no they don’t extend your battery life. Both Android and iOS are fairly sophisticated about managing apps in the background.

Turning off WiFi, GPS and Bluetooth will not extend your battery life, at least not meaningfully. These radios use such little power that turning them off will not give you any noticeable advantage. The fact is that your cellular radio signal (e.g., LTE) and your display (specifically when the screen is on) are the two primary consumers of battery life — and turning these off render your mobile device somewhat useless.

Lastly is the question of “should I charge the battery to 100%?” Well, yes! but you don’t have to if you don’t want to or can’t. In other words, stop thinking about it. The battery is fine whether you charge it to 100% or to 80% or anything else. Sure, for those of you who are battery geeks, yes, you will get more cycle life if the battery is not charged to 100%. But to the average population, you can do whatever you like — your usage is not wrong. These are design specifications that the device manufacturer is thinking about on your behalf.


Yes, as long as the memory of a 95-year old suffering from Alzheimers!! Sarcasm aside, lithium ion batteries have zero memory effects. Now, if you are a techie intent on confusing your smartphone or mobile device, here’s a little trick. Keep your device’s battery between 30% and 70% always….this will confuse the “fuel gauge,” that little battery monitor that tells you how much juice you have left. The battery will be just fine but the fuel gauge will not report accurately. Every so often, the fuel gauge needs to hit close to zero and 100% to know what these levels truly are, otherwise the fuel gauge will not accurately report the amount of battery percentage. This is like your gas gauge in your car going kaput…it does not mean that the battery has memory or other deficiencies. Should you suspect that your fuel gauge is confused, charge your phone to 100% and discharge it down to 10% a few times. That is sufficient to recalibrate the gauge.


This one garners a lot of media interest. Every time a research lab makes a new discovery, it is headline news and makes prime time TV. The reality is that the path from discovery in the lab to commercial deployment is extremely rocky. There have been dozens such discoveries in the past 5 – 10 years, yet virtually none have made it into wide commercial deployment. History tells us it takes over $1 billion and about 10 years for a new material to begin its slow commercial adoption cycle….and for now, the pipeline is rather thin. Additionally, present lithium ion batteries continue to improve. Granted, it is not very fast progress, but there is progress that is sufficient to make great products….just think that current battery technology is powering some great electric vehicles.

Let me be more specific. Present-day lithium ion batteries are achieving over 600 Wh/l in energy density — that is nearly 10x what lead acid batteries can deliver. This is enough to put 3,000 mAh in your smartphone (sufficient for a full day of use), and 60  kWh in your electrical car (enough for 200 – 250 miles of driving range). With the proper control systems and intelligence, a mobile device battery can last 2 years or more, and an electric vehicle battery can last 10 years. Does it mean we stop here? of course not, but this sense of urgency to develop new materials or chemistries is rather misplaced. Instead, we need to keep optimizing the present batteries materials and chemistries. Just reflect on how silicon as a semiconductor material was challenged by other candidate materials in the 1980s and 1990s (do you remember Gallium Arsenide), only for it to continue its steady progress and become an amazing material platform for modern computation and communication.


What made silicon the king of semiconductor materials is its amazing cost curve, i.e., decreasing cost per performance, aka Moore’s law. Now, lithium ion batteries don’t have an equivalent to Moore’s law, but, the cost of making lithium ion batteries is dropping fast to the point they are rapidly becoming commoditized. A battery for a smartphone costs the device OEM somewhere between $1.50 and $3.00, hardly a limiting factor for making great mobile devices. GM and Tesla Motors have widely advertised that their battery manufacturing costs are approaching $100 /kWh. In other words, a battery with sufficient capacity to drive 200 miles (i.e., 50 to 60 kWh) has a manufacturing cost of $5,000 to $6,000 (excluding the electronics)…with continued room for further cost reduction. It’s not yet ready to compete with inexpensive cars with gas engines, but it sure is very competitive with mid-range luxury vehicles. If you are in the market for a BMW 3-series or equivalent, I bet you are keeping an eye on the new Tesla Model 3. Tesla Motors pre-sold nearly 200,000 Model 3 electric vehicles in the 24 hours after its announcement.  This performance at a competitive price is what makes the present lithium ion batteries (with their present materials) attractive and dominant especially vis-a-vis potentially promising or threatening new chemistries or new materials.


Why do we not worry about the immense flammability of gasoline in our vehicles? Isn’t combustion the most essential mechanism of gas-driven cars? Yet, we feel very safe in these cars. Car fires are seldom headline news. That’s because the safety of traditional combustion engine cars has evolved immensely in the past decades. For example, gas tanks are insulated and protected in the event of a car crash.

Yes, lithium is flammable under certain but well known conditions. But the safety of lithium ion batteries can be observed as religiously as car makers observe the safety of combustion engines. It is quite likely that some isolated accidents or battery recalls may occur in the future as lithium ion batteries are deployed even wider than they are today. In mobile devices, the track record on safety has been very good, certainly since the battery industry had to manage the safety recalls at the turn of the century. Is there room for progress and can we achieve an exceptional safety record with lithium ion batteries? Absolutely yes. There are no inherent reasons why it cannot be achieved, albeit it will take time, just like the automotive and airline industries have continuously improved the safety of their products.

06Mar 2016

General Motors announced in January the all new Chevy Bolt, and Tesla Motors followed recently in pre-announcing the long anticipated Model 3. Both cars are new and novel, the first electric vehicles ever to boast of a driving range of 200 miles at a price before incentives between $35,000 and $40,000.  My 2013 Ford Focus Electric with its now degraded 70-mile range feels all too aged. It only goes to speak of the rapid transformational pace of the automotive industry, and it is only the beginning.

But even as an enthusiastic advocate and proponent of pure electric vehicles, also known as battery electric vehicles (BEV), I cannot ignore the incredible evolution that hybrid vehicles (HEV) have witnessed since the introduction of the first Toyota Prius in 1997. New plug-in hybrid models now span nearly every major auto manufacturer and include the Toyota Prius, the Chevy Volt, and other models such as the BMW X5 xDrive plug-in. Hybrids are no longer considered cars from the future….their sale volumes rocketed when gasoline prices soared a few years ago, but the presently low gas prices are making average consumers pause and ask whether being green also means keeping green in your wallet.

So let’s do a simple exercise in this post and compare the actual cost of maintaining three types of vehicles. The first is a conventional car with an internal combustion engine (ICE), assuming it has an average efficiency of 25 mpg (9.4 L/100km). The second is a hybrid vehicle (HEV) similar to the Prius. I assume an average efficiency of 50 mpg (4.7 L/100km). The last type is a pure electric (BEV). Based on my driving experience and other published data, I assume an average efficiency of 270 Wh/mi, or using the US EPA unit for energy consumption, it works out to 106 MPGe (equivalent mpg). I further assume that the average cost of electricity is $0.10 /kWh — that’s roughly what I pay when I charge my car at night, recognizing that in some geographies outside of our home state of California, that cost may be quite lower. Finally, for the benefit of this exercise, I ignore traditional maintenance costs such as oil change — I am happy to confirm that I have spent zero on oil changes for my Ford Focus Electric, precisely as the Ford dealership claimed !

I will spare you the math, but the cost of driving your vehicles (in $ per mile driven) depends largely on the prevailing price of gasoline.


When gas peaked at $5.00 per gallon several years back, HEV and BEVs provided a significant improvement over conventional cars when it came to cost per mile (again, I am giving ICE a benefit of zero maintenance cost which is not true in real life). But at today’s gas prices hovering near $2.00 per gallon, the financial advantage of HEVs and BEVs shrinks substantially. Furthermore, the advantage of BEVs diminishes substantially when compared to HEVs. At $2.00/gal, HEVs are only a mere additional penny per mile to operate…so now one can understand why cabbies prefer their HEVs over the old Ford Victorias. Additionally, given that HEVs don’t suffer from “range anxiety,” more cars like the Chevy Bolt and the Tesla Model 3 need to hit the showrooms before they can be seriously considered as primary household vehicles (as opposed to commuting second cars).

Both limited range and low gas prices remain strikes against the wide adoption of BEVs. So long gasoline prices stay in the foreseeable future near their all-time lows, government incentives and regulations will continue to play an important role in the promotion of BEVs — these incentives include the highly coveted white stickers to use the carpool lane in California. It will take a serious increase in gas prices before incentives for BEVs come into question. This, however, is a chicken-and-egg problem. With the United States well on its way to energy independence, it is difficult to see scenarios in the near future that will catapult the price of gasoline…unless one begins to consider gloom-and-doom war scenarios, which I will leave to others to consider.

11Jan 2016

A lithium-ion battery cell for a smartphone costs the device OEM somewhere between $2 to $4 depending on its capacity and other design attributes. It constitutes about 1 to 2% of the entire cost of the mobile device. In contrast, a lithium-ion battery for an electric vehicle can range between $7,000 and $20,000, making it by far the most expensive item in the cost of the vehicle. So what drives these cost figures?

The primary cost metric of a lithium-ion battery (not including the pack electronics) is dollars per kilowatt-hours, abbreviated as $/kWh. That’s the cost of each unit of energy measured in kWh. A small smartphone lithium-ion battery stores about 10 Wh, or 0.01 kWh. A Nissan Leaf has a battery capacity of 24 kWh; the Tesla Model S can reach up to 85 kWh.

Today’s metric stands near $200 /kWh (or $0.20 /Wh) for consumer-grade batteries, and the cost continues to decline. See this earlier post to learn how these costs declined by 10X from 1995 to 2015, primarily as a result of increasing installed battery manufacturing capacity. Cost of lithium-ion batteries for electric cars is also declining…recent announcements from General Motors suggest a cost of $145 /kWh for their EVs declining to $100 /kWh in 2021. Now that’s GM’s numbers and they don’t necessarily reflect the cost structure for all EV makers — albeit it is highly rumored that Tesla’s figures are in the same vicinity. At $145 /kWh, the estimated cost of the lithium-ion battery of the newly announced Chevy Bolt is $8,700, plus an additional $3,000 for the pack electronics, for a total of nearly $12,000 — not too bad for an electric car that is advertised to go 200 miles and priced at about $40,000 before government incentives.

While the cost structure of manufacturing batteries is usually confidential to the battery makers and their large customers, there is sufficient information from various market reports and intelligence that provide insight into the components that make up the total cost figure. The chart below breaks down the cost for consumer-grade batteries into three basic categories: i) material costs, ii) labor costs and iii) overhead costs and profits.

Cost components of a lithium-ion battery

Cost components of a lithium-ion battery

The material costs are by far the largest contributors — about 60% of the total cost. For lithium-ion batteries made using lithium-cobalt oxide cathodes (LCO, used in consumer devices) or nickel-cobalt-aluminum cathodes (NCA, used in Tesla), the price of raw cobalt is a major component, presently priced at $10.88 per lb. That translates to nearly $10 to $15 /kWh just for the cost of raw material for the cathode — before processing and manufacturing. If you are wondering about cobalt sources, mines in Africa are the largest production sites. The good news are that cobalt pricing is at its lowest since it hit a peak of $50 per lb in 2007, and the US Department of Energy does not deem its supply at risk.

Labor is a relatively minuscule component of the overall cost — battery manufacturing combines significant automation in countries with low labor costs, in particular China. Overhead costs are, however, much more significant at 30% — they include depreciation of the capital, energy costs, R&D, sales and administrative…etc. As I have mentioned in prior posts, profit margins tend to be in the single digit…the financials of a battery manufacturing business are nothing to write home about.

This situation and increasing dominance of China in battery manufacturing has led many companies to be material suppliers. For example, the cathode material market is nearly $2.5B dominated by large conglomerates such as Tanaka and Mitsubishi in Japan and 3M in the US. Another example is the electrolyte market: at nearly $1B, it is dominated by the Japanese companies Stella Chemifa and Kanto Denka Kogyo.

Looking at the chart above gives an idea on where future cost-reduction measures are likely to be. First, reducing overhead costs…one can think of Tesla’s Gigafactory in Nevada as an example. Second, reducing the material costs. This is a longer term process with larger players, e.g., the large automotive manufacturers and to some extent the large consumer manufacturers such as Apple, exerting influence on their supply chain to improve efficiencies and reduce costs. The result: A goal of $100 /kWh by or before 2020 is within reach. This is widely accepted as the cost point where EVs can become affordable to larger segments of society and compete effectively against traditional combustion-engine vehicles.

However, at these cost levels, it becomes really hard to justify amortizing large R&D costs that may be necessary for major future battery breakthroughs. Instead, what is more likely to happen is a gradual evolution of battery materials and designs funded by modest R&D investments, and consequently continued commoditization of the lithium ion battery. This is not a bad thing to happen.

17Dec 2015

The California Department of Motor Vehicles (DMV) proposed today a new set of rules that will govern the operation of autonomous vehicles on the state’s roads. Simultaneously, Google announced that it was spinning off its self-driving car unit into a standalone business setting up the stage for a fleet of self-driving taxis that will compete head-to-head with Uber, which itself is investing heavily in self-driving vehicles. Tesla, GM, Mercedes Benz, Ford and several others have not been shy in the media, all announcing efforts and prototypes towards autonomous driving.

The Google (or perhaps more aptly under its new name of Alphabet) pod-like self-driving vehicles are powered by lithium ion batteries, so it is fair to say that it is only a matter of time before electric powertrains become the foundation for this new vision of autonomous cars. The race is early, still very early, but the stakes are potentially immense as several players pursue their vision for autonomous electric vehicles.

One of the early metrics of a race worth gauging is each player’s present IP position. It is not a simple question to answer but one can glean some insight — autonomous electric vehicles are very sophisticated systems using complex components, so one would expect that intellectual property will play a central role both in the development of this market segment as well as eventual litigation among the participants.

The next two charts show the extent of the present IP position for a select number of companies. The chart on the left shows the number of US patents issued since 2000 covering two categories: i) battery technology including battery battery materials, manufacturing and battery management systems, and ii) technologies related to designing and building electric vehicles. For the time being, I will focus this post on the “electric” portion of this race, addressing the battery and electric systems for these vehicles — leaving autonomous driving for others to discuss. I assume that the number of issued US patents will reflect within reason the amount of know-how a company possesses in battery and EV technologies.


 Patent positions related to autonomous electric vehicles


The first observation that stands out is the large number of US patents that Toyota has secured in both areas of batteries and electric vehicle systems. They eclipse the number of patents issued to Tesla and GM. Honda and Ford, two companies that have been relatively quiet in the media, are clearly building their foundations. The German automakers, judging from their US patent portfolio, seem to be lagging — though this should not be misinterpreted as losing or lagging in the race. Apple has not yet publicly announced that it plans to build cars, but rumors abound in this respect and as such, the companies is categorized with the automakers. Their portfolio, however, is heavily biased towards battery technology, courtesy of their prowess in consumer devices.

Automakers rely heavily on suppliers of components and subsystems. Among the well-established ones, Robert Bosch stands out with a sizable bag of issued US patents covering both batteries and electrical vehicles. Samsung and LG, two large suppliers of electronic components to the Korean car makers, have a strong IP position in building batteries owing to their respective battery divisions, SDI and LG Chem — but there is not much evidence of strong IP in electrical powertrain and electric vehicles. It is also surprising to see Delphi and Johnson Controls lagging in both categories — could this mean that the automakers are choosing to own and control key technologies instead of outsourcing them to their traditional supply chain partners? Time will tell.

In this evolving race and ambitious vision for the future, these statistics are merely just perturbations for the time being. However, given enough time, they could amplify and influence the outcome of who will win and who will not. Stay tuned!

10Dec 2015

Before I start this post, I encourage all new readers to go back to the early posts if they desire to learn more about the basics of lithium-ion batteries and their operation.

This post is dedicated to deciphering the growing complexity of charging stations for plug-in hybrid and pure electric vehicles (xEVs), especially as their popularity grows among drivers. If you own an electric vehicle, a Tesla, a Nissan Leaf, or other models, chances are you are using one or more of these charging methodologies…and if you so desire to fast charge your Tesla or electric vehicle, this post will give you the insight to know what type of charging station you may need.

The charging of xEVs, whether at your home or at dedicated charging stations, is usually governed by a set of standards agreed upon by a vast number of contributing organizations, such as the automakers, electric utilities, component makers and many others. Several organizations including SAE International, ANSI and IEEE have led the  coordination and development of such standards — they are numerous — covering, for instance, the types and interoperability of connector designs and charging power levels (SAE J1772), communication and signaling protocols  (SAE J2931/1) between the xEV and the charging station (also known as EV supply equipment — EVSE), wireless charging (SAE J2954) as well as many others related to diagnostics, safety, DC charging…etc. Today’s post addresses charging from the perspective of the SAE J1772 standard and its competing standard CHAdeMO.

So let’s start with understanding what charging levels are and how they are defined by the various standards. There are 4 levels of charging:  two levels using conventional AC charging, and two additional levels using higher-voltage DC  charging. They are summarized in the next table:

Fast charging a Tesla or an electric vehicle

Let’s start with AC levels 1 and 2. Level 1 is what you get using the charging cord supplied to you by the car maker if you own an xEV. It plugs into the standard 120V household outlet and delivers, in theory 1.7kW. Some of you are probably tempted to multiply 120V by 20A and realize that’s more than 1.7kW…if you are that person, remember that this is an AC current, so you need to multiply by 0.707. This is the maximum power delivered to the connector plugged into the car. It is not the power delivered to the battery. The battery’s power is delivered by an on-board battery management system (on-board charger) that has to convert the voltage to a level appropriate for the battery. In reality, the car battery receives a best-case power of about 1.2 – 1.3kW to account for the electrical efficiency of the system — taking an average consumption of 250Wh per mile, that is equivalent to about 5 to 6 miles for each hour of charging…ouch!

Reality is a little worse than that: standard household power plugs are limited to 15A (instead of 20A) thereby decreasing the power delivered to the battery to a measly 900W. At this power level, a Nissan Leaf’s battery (nominal 24kWh / effective 20kWh) takes 22 hours to fully charge. Yikes! Now one can begin to understand why xEV owners do not line up near a Level 1 charger…but it does get crowded at a Level 2 charger.

AC Level 2 uses a 240V single-phase mains. The lowest current level is 20A corresponding to a maximum power (again at the output of the connector) of 3.4kW. The typical public charging stations, such as the ones managed by ChargePoint, provide 6.7kW. However, there is a catch. The on-board charging circuitry in your xEV must be able to use that power. Early Nissan Leaf models had 3.3kW-circuitry — in other words, regardless of the maximum power at the charging station, the maximum power the car is willing to accept is 3.3kW.  Newer xEV models, e.g., Nissan Leaf, Ford Focus Electric, Chevy Volt, have on-board chargers capable of up to 6.6kW. Again, this means if the charging station were to provide 19.2kW, your car cannot accept more than 6.6kW…this is by far the most common charging level as dictated by the presently deployed infrastructure. It equates to about 25 miles for each hour of charging. Once again, using the Nissan Leaf as an example, its battery will fully charge in 3.5 hours with a Level 2 charger (6.6kW). That’s not fast charging, but it sure is a heck-of-a-lot faster than Level 1.

Fast charging with DC gets more complex because there are competing approaches. Of course, we are also now talking about insanely high power levels, and consequently very expensive charging stations  and associated installation costs ($50,000 to $100,000 each).

SAE has the J1772 Combo DC standard. CHAdeMO has another competing standard. Tesla has its own proprietary fast-charging using their network of Superchargers (though not DC). But what they have in common is that they all seek to provide high power levels to the vehicle…up to 120kW. This infrastructure is still relatively scarce — Tesla is the only one aggressively deploying fast charging Superchargers along specially designated highway corridors.

Naturally, charging a car battery at such high power levels begs a new series of questions on whether this creates any significant and permanent damage to the battery. The brief answer is: YES, damage does occur…but super fast charging is so rare that no one is really paying attention to this question, at least not for now. Besides, with the exception of Tesla, your average xEV cannot charge faster than 6.7kW, so having a fast charging station is a moot point.

A final word on fast charging the Tesla batteries. At 120kW of input power, this equates to a charging rate of 120/85 = 1.4C — this is guaranteed to cause serious damage to your Tesla battery if you were to charge on a regular basis. But then again, if Elon Musk and Tesla Motors are willing to cover you under their warranty, do you really care?