Blog Post

The battery performance deficit disorder

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Batteries fail — its ‘s certain as death and taxes. Rechargeable batteries at least offer the possibility of repeating the cycle. But alas, the story cannot repeat indefinitely. One cheerful thought after the other, yes?

But wait, there’s more . . . Add to their inevitable demise an overall lackluster performance in battery storage technology, and we have ourselves the makings of a blog post on the failure of batteries to live up to their promises.

To set the stage, the specific energy of gasoline — measured in kWh per kg, for instance — is about 400 times higher than that of a lead-acid battery, and about 200 times better than the Lithium-ion battery in the Chevrolet Volt. We should not expect batteries to rival the energy density delivered by our beloved fossil fuels — ever.

recent article in APS News reported on an emerging view that batteries are failing to live up to our dreams in the electric car realm:

Despite their many potential advantages, all-electric vehicles will not replace the standard American family car in the foreseeable future. This was the perhaps reluctant consensus at a recent symposium focused on battery research.

I was somewhat stunned to see this article. I am accustomed to seeing articles emphasizing the possible — albeit often improbable, in my mind. Also appearing in the article is a quote from Paul Alivisatos, an accomplished physicist, summarizing the need for further research:

“It remains true today, as in the past, that we need a fundamental understanding of the physics of how energy-conversion processes take place, at a much deeper level, in order to achieve a truly sustainable energy future.”

Rephrasing: the physics we currently understand is not sufficient to deliver the kind of battery we need to make the future work without fossil fuels. Red flags go up for me when it is our understanding of physics rather than practical engineering challenges standing in the way — as serious as the latter can be. Physics limitations instantly present a much taller order to overcome.


I’m sure everyone has tales of how batteries have let them down — ranging from the merely annoying to life-threatening situations. I find that I am more often disappointed than pleasantly surprised when it comes to batteries. Here are some examples:

  • I frequently go for months without driving my truck. The battery is often dead when I try to start it. Lead-acid batteries only get worse if left in a discharged state, so it’s a runaway process. Fortunately, I live on a hill and can often roll-start my way back onto the road.
  • The rechargeable NiMh batteries I use for small electronics devices are rated for 1000 charge cycles. I’ll bet I only get about 15–20 cycles before noticing a serious degradation in performance.
  • The first set of lead-acid batteries I used with my home-built solar photovoltaic system only lasted two years before showing substantially reduced capacity. A newer set is still in good shape after 2.5 years, but the drop in performance can be pretty fast, I have found.
  • Lead-acid batteries for cars tend to last 5–6 years, often failing with little warning, in many cases resulting in being stranded.
  • New laptop batteries seldom fail to delight their owners in how much longer the charge lasts compared to the previous generation batteries. But give it a few years and it is not uncommon to be operating at half the original capacity.
  • Batteries left in a device for a long time can develop corrosive crud around the terminals, often in hard-to-clean places.

A counter-example is the occasional amazement I experience when alkaline batteries in a device that has not been utilized in years crackle to life after all that time — if the batteries haven’t gooped themselves up, that is.

Energy-Power Tradeoff

The chief measure of a battery, in my mind, is how much energy it can store. But it makes sense to adjust this concept to the size or mass of a battery. Obviously, a more massive and voluminous battery can pack in more energy. So for a given mass (we’ll take a kilogram), we want to know how much energy a battery can store, called specific energy.

At low power demand (sipping rather than gulping), lead-acid batteries tend to hold about 30–40 Wh per kilogram (one Watt-hour is equivalent to 3600 J, or 0.001 kWh of energy). Ni-MH batteries score 45–60 Wh/kg, and Lithium-ion gets about 120–180 Wh/kg. Part of the reason for Li-ion’s better performance is that lithium itself is lightweight; by volume lead-acid has about 40percent the capacity of Li-ion. Gasoline, at 36.6 kWh/gal, has a specific energy of 13,800 Wh/kg. Off the charts!

As power demand increases, the battery flags, and will not offer as much total energy. Obviously, the battery discharges faster under heavier power demand, but the effect is exacerbated by less actual energy available. This is best shown on a Ragone plot, in which specific energy is plotted against specific power.

Note that the Internal Combustion Engine (ICE) exceeds both specific energy and power goals for vehicles (the mass must include engine weight, rather than the fuel by itself). Fuel cells provide decent specific energy, but typically insufficient power (per kilogram). Capacitors, including super-capacitors, discharge super-fast with lots of power, but have very low specific energy.

As useful as this plot is, it does not convey the whole story. While it looks like Li-ion meets the the goal for plug-in hybrid electric vehicles, this does not necessarily remain true if demanding 5,000 deep charge cycles, a ten-year lifetime, a moderately inexpensive product, etc.

Spider Diagrams

The U.S. Department of Energy teamed up with the automotive and battery industries to define benchmark performance targets for batteries that would result in electric vehicles being competitive with ICE vehicles on a mass-produced basis. The resulting coalition was called USABC/FreedomCAR, and their various target requirements are available here, with a useful summary presentation also available. Below is a subset of the target parameters pulled from these sources, and I have also thrown in the Chevrolet Volt for a side-by-side comparison to current capabilities.

The 300 mile (580 km) range for the pure electric vehicle (EV) comes from the presentation rather than the official USABC source, and does not look right to me based on the 40 kWh battery size. Electric cars typically need 30 kWh of storage for each 100 miles of driving (about what the Volt, Leaf, and Tesla achieve, based almost entirely on air resistance — not battery technology). So I would expect the 40 kWh battery pack associated with the EV goal to deliver half as much range as what’s in the table.

Some of the figures for the Volt deserve explanation, since many cannot be directly looked up, and require inference and calculation. Firstly, the 2013 model battery pack has a capacity of 16 kWh, but only 10.5 kWh are made available so-as to avoid potentially damaging deep discharges. Meanwhile, I have no choice but to use the entire battery pack mass and volume (197 kg; 100 L) in conjunction with the partial 10.5 kWh charge in calculating energy densities, because available energy density is what’s important.

For lifetime and cycle computations, I use the 100,000 mile, 8-year guarantee on the battery, together with the estimated 37 miles per gallon (MPG) on gas alone and 98 MPG for combined gas/electric. This implies an expectation that about 62,000 of the 100,000 miles will be driven under battery power. If recharges typically happen after 30 of the 38 miles are spent (corresponding to 80percent of available capacity), this translates to about 2,000 deep cycles. Perhaps this is pessimistic in the sense that most guarantees correspond to aminimum expected performance. But offsetting this is the fact that the USABC targets are specified for end-of-life performance, whereas I use the beginning-of-life numbers for the Volt. General Motors estimates a 10–30 percent degradation at the end of 8 years (100,000 miles).

A comparison between actual performance and target performance can be cleverly displayed graphically in a “spider chart,” as illustrated below for the plug-in hybrid performance as of May 2011 (I first saw such diagrams in a presentation by Venkat Srinivasan, in 2008).

We can make our own spider diagram for the Volt, based on the numbers in the table. Please excuse the sub-optimal placement of labels, etc.

Besides looking like some sort of cool fighter jet in a dive, the diagram highlights performance deficits on several fronts. It is not terribly hard to get lots of current out of a battery, translating to more-than-adequate power performance. But all other measures fall short of the goals by varying degrees. The APS article intones that we should not hold our collective breaths to see a march of progress in lithium-ion technology at a level that would satisfy this (still hungry) spider. In practice, improving one aspect of performance tends to decrease another somewhere else (see the piece by Srinivasan for more on this principle). So it’s not a simple matter of advancing on all fronts independently and incrementally.

Full Cost of Electric Drive

Let’s say you pay $0.10 per kWh for electricity delivered to your home. Charging the Volt battery with 10.5 kWh at 90 percent efficiency to replace the drain from 38 miles of driving will cost $1.17. If using gasoline alone, the same car uses about a gallon of gas to go the same distance. Let’s put the cost of that gallon at $4.00. Electric looks pretty good, at these rates!

Now figure in the estimated price of the Volt battery at $8,000 (a disputed number, but GM has not revealed the actual cost). If we get 62,000 miles of electric drive out of the battery, we will spend $1950 on electricity for charging, plus $8000 for the battery. That’s $9,950. The same distance on gasoline would cost $6500. Not an order-of-magnitude difference, but still gasoline currently wins.

If the price of gasoline goes up (it will; but so will electricity), and the cost of the battery goes down (it should), the two may cross. But there are other added costs to the Volt (or hybrids in general) besides just the battery. After all, hybrids can’t jettison the ICE, and require an electric drive train to boot. Even the fact that the space occupied by the battery forces bucket seats in the back of the Volt is a “cost” that must be paid.

Beyond Cars

Batteries are, of course, useful for purposes other than transportation. While transportation hardship may be the most pressing problem in the decades following peak petroleum production, solar and wind resources cannot scale to be very large without a viable storage solution.

I worked out in an earlier post how large a lead-acid battery would have to be to support the entire U.S. energy demand in the presence of solar/wind intermittency. It turned out that our estimates for recoverable lead in the world do not satisfy the need. Lithium and Nickel are even more constrained. It is possible that some other approach like sodium-sulfer or zinc-air can step in. But these are already relatively well-known options and have not blazed a wide path into storage over the past few decades.


Don’t get me wrong: even though I dwell on the shortcomings of batteries in this post, I still hold a net positive view. When it’s dark at my house, my refrigerator, television, computers, and internet goodies are all powered by stored sunlight in lead-acid batteries. My laptop battery gets me through many a bus ride and an occasional airplane ride. Batteries really do work, and provide value. Moreover, electric cars are more than a notion or fantasy: they are actually on the road getting people where they want to go.  Despite their lackluster performance next to fossil fuel storage, batteries still beat the pants off of mechanical or gravitational storage.

And even though I might appear to be picking on the Chevy Volt by highlighting its deficiencies, I actually rather like the design point (electric vs. gasoline range hits the sweet spot, in my view). In fact, I was half way to buying one. By half way, I mean that if the price were cut in half, I would surely have one now.

The real point is that batteries fall pathetically short of our customary fossil fuel energy storage medium. When we wake up to a declining global availability of petroleum, we won’t just switch over to electric cars. We may not be able to collectively afford such a transition, given the huge up-front costs in both money and energy. Where will the prosperity come from? If oil shortages drive recession in the usual fashion, expensive options may be off the table.


The same author of the APS article referenced above wrote an extended version, worth a look.

This post originally appeared on Tom Murphy’s blog, Do the Math: Using physics and estimation to assess energy, growth, options.

Tom Murphy is an associate professor of physics at the University of California, San Diego. An amateur astronomer in high school, physics major at Georgia Tech, and Ph.D. student in physics at Caltech, Murphy has spent decades reveling in the study of astrophysics. He currently leads a project to test general relativity by bouncing laser pulses off the reflectors left on the moon by the Apollo astronauts, achieving one-millimeter-range precision. 

22 Responses to “The battery performance deficit disorder”

  1. Weekendmoe

    Hillarious article :)
    I stopped reading after reading: “The rechargeable NiMh batteries I use for small electronics devices are rated for 1000 charge cycles. I’ll bet I only get about 15–20 cycles before noticing a serious degradation in performance.”

    My laptop as well as the Model S uses Lithium-ion. After approximately 400 charges the laptop still have close to original battery performance.
    Norwegian Nissan Leaf owners report zero battery reduction after 40-50.000 miles.

  2. When bringing the Tesla Model S into the conversation, it’s important to compare apples to apples. It’s a luxury performance vehicle, and its cost of ownership should be compared to a similar ICE powered vehicle, such as an Audi S6, BMW M5 or Mercedes AMG. Fuel efficiency is not a primary concern for owners in this vehicle category, but maintenance is a concern – especially when the warranty has expired. Anyone who has owned an out-of-warranty Audi or BMW can relate to this.

    The Tesla Model S has less than ten percent as many drivetrain parts as ICE powered cars, so there’s a lot less to break or wear out. Hybrids that include an ICE do not reap this benefit. The regen-braking on the Model S is expected to result in up to five times more brake component life. Time will tell, but aside from eventual battery replacement, it appears that the Model S will be a less expensive way to enjoy luxury and performance. The Model S is not being built for the average family, and neither was the flat screen television when it cost $10,000. But brave new technology has to start somewhere. Within five years, Tesla will have a $30,000 car for mass consumption that benefits from the lessons learned with the Model S.

    Only five years ago a car like the Model S would not have been possible. The power-to-weight ratio of the battery pack was not there yet. With continued improvements in Lithium battery technology, EVs might indeed become more than outliers.

  3. Quote:

    “the specific energy of gasoline […] is […] about 200 times better than the Lithium-ion battery in the Chevrolet Volt. We should not expect batteries to rival the energy density delivered by our beloved fossil fuels — ever.”

    IC engines have about 25% effish, large modern electric motors more than 90%. Also, the latest generation ENVIA Lipo cells have about 400Wh/kg. Taking that into consideration the specific energy density is not nearly 200 times better, a fraction of that at best. Maybe 30 or fourty times, if you are generous and don´t count the massive amount of electricity that go into the refining and transportation of gasoline.

  4. Hmm… Well, a proper conversion should yield 4 miles per lead-acid battery, and my Vw beetle holds 120 pounds of gas. Assuming a lead acid battery weighs 8 pounds, I can fit 15 in my beetle. That means–for the same weight–I would get 60 miles per fill-up. On gas, my car has about a 450 mile range.

    Tell me if I’m wrong, but 60×400(times energy density)=a whole lot more than 450…

    Perhaps we need to remember the efficiency of battery to energy is insanely better than gas to energy?

  5. Douglas Hvistendahl

    For cities and high density suburbs, electrically powered personal mass transit should be looked at. For a possible example, look at the J-pod. Look ma, no batteries!

  6. The entire energy cycle is completely ignored here. Transport losses and conversion losses show a real advantage for electric vehicles vs gasoline. Plus, I don’t know of anyone that regularly drives over 50 miles per day unless they drive over-the-road trucks.

  7. Bob Wallace

    Here’s where I think your analysis breaks down – the assumption that we need the range provided by current oil powered vehicles.

    First, most people will charge at night when their car is parked. With wireless charging it will be as painless as parking over the “charging turtle” in your garage/parking space/along the sidewalk. People will wake up with a full charge as opposed to driving down to a quarter tank/whatever and taking time to go to the gas station.

    Second, most people rarely take long trips in their cars. They might do a full 500 mile drive a few times a year. Those 500 mile drives are generally broken by at least one stop for gas and another for food. An EV with a 180 mile range and ~20 minute, 90% charging will allow for a 500 mile drive day with two modest length stops. About the same as driving with gas but for about 1/4th the per mile cost.

    The Toshiba SCiB lithium-ion battery used in the Honda FiT EV and Mitsubishi i-MiEV will recharge 95% in less than 20 minutes. It is rated for 4,000 cycles. Put that battery in a 180 mile EV and you’ve got a 720,000 mile battery.

  8. Great article. The one thing that too rarely gets mentioned when comparing customary fossil fuel to electric is refuel/recharge time. Unless the industry can come up with a scalable battery swap infrastructure, the 3-5 minutes to refill a tank compared to hours for electric (assuming charging stations are as plentiful as gas stations) is going to be as big a hurdle as the capacity issue.

    • Bill Brown

      You’re right there, Mike, assuming one’s trips require recharging while “out”. The overwhelming majority, if not close to 100%, of BEV buyers perform their entire daily commute with the vehicle on a single charge, that is their predictable daily distance traveled is less than the range of a complete charge done at home. I don’t foresee anytime soon that drivers will be pulling over every hour and a half to charge for 4 hours at the next available 240V charger. Those who daily exceed the single charge range or who frequently take long trips would be silly to rely on a BEV only.

  9. Mark Renburke

    Mark Renburke
    Great info and analysis! Conservative, I must say though (in the non-political meaning:)

    I’d like to present a real world scenerio with the Volt using your methods. I’m on track do 95% battery-electric miles (currently 6281/6578 or 95.5% EV). So in my case this would result in a gas savings of easily over $7000. Here are the calculations:

    Electric miles:
    95k miles/38 miles per charge = 2500 charge cycles * $1.17 = $2925

    where as gas miles would cost:
    95k miles/37 mpg = 2568 gallons * $4.00 per gallon = $10,270

    $10,270 – $2,925 = $7,345

    So I will save and estimated $7,345 in gas for 95k miles. In reality I will save even more as I charge at work so my employer pays for half my commute, and I use a variety of free public charging points.

    This doesn’t even touch on all the other benefits of driving an EV+H and the Volt in particular – topic for another, longer discussion elswhere!

    • Lee Conley

      Totally agree Mark. Then there is oil, oil changes, and an engine with magnitudes of complexity over an electric system, leading to further repairs. The horse was far cheaper more reliable than when the car came out. Lets give E a chance. The one thing that really burns me though, is the arbitrary fuel prices implemented by the oil companies, while being subsidized by our government, and protected by our military. WTF? Talk about being manipulated. At this point I would rather pay more for an eCar up front, and perhaps even invest in some solar panels, than to continue to be dictated to by big oil. Then theres the environment… drill baby spill! No thanks, my next car will be electric.


  10. The author has an agenda that is poorly hidden. The article is full of suggestions and speculations using key words as should, may, while, but, some, etc., using rhetorical questions and first person references. He seems to conveniently miss the fact that although gasoline may have more energy per mass/volume, only a very small part of that energy can be converted to kinetic energy, the rest is wasted as heat. He considers the batteries in the Volt that are about 4 years old in terms of technology. He lists the 2013 Volt battery capacity as 16 kWh (it is 16.5), the available capacity as 10.5 kWh (it is 10.8), but fails to disclose why GM made the decision to reserve such a large margin, or include a thermal management system. Picking specifically the Volt as an example of an electric car rings the alarm. Why not consider the Leaf, or the MiEV, or the Focus EV, or the Fit EV, or the Model S?

    • loopyduck

      This is amusing because *your* agenda is pretty naked, yet *also full of errors and ignorance*

      “He seems to conveniently miss the fact that although gasoline may have more energy per mass/volume, only a very small part of that energy can be converted to kinetic energy, the rest is wasted as heat.”
      And you seem to conveniently have forgotten how to do basic arithmetic. On average, about 1/5 of the energy in gasoline makes it to pushing the car. So that’s still 40 times the energy density compared to li-ion. And this completely ignores various schemes being studied to allow mass-produced cars to recover thermal energy from the cooling and exhaust system, of which there is plenty as you yourself pointed out.

      “He considers the batteries in the Volt that are about 4 years old in terms of technology.”
      Pray tell, would you mind pointing out a particularly major advance in battery technology coming to the market in the upcoming model year? The fact of the matter is, every battery breakthrough is always five years from market, at which point it’s figured out that it’s too fragile, not mass-producible, very expensive, or all of the above.

      “He lists the 2013 Volt battery capacity as 16 kWh (it is 16.5), the available capacity as 10.5 kWh (it is 10.8), but fails to disclose why GM made the decision to reserve such a large margin, or include a thermal management system.”
      One need only look at the case of the so-called “bricked” Tesla Model Ss to answer your question. Should you leave the Volt at home for a few weeks to go on your vacation, having an additional capacity acting as a buffer helps avoid the need for a costly battery replacement. Also, as you know (or SHOULD, given your attitude), batteries degrade over time. Given the same degradation rate, which do you think would impact driving performance sooner, a 12 kWh battery with 10.8 kWh available or a 16.8 kWh battery with 10.8 kWh available? And while you think this is some sort of anomaly, take note: the batteries in Toyota Priuses only discharge to 50% of total capacity, and only charge up to 80%!

      “Picking specifically the Volt as an example of an electric car rings the alarm. Why not consider the Leaf, or the MiEV, or the Focus EV, or the Fit EV, or the Model S?”
      Because with the exception of the Model S, none of the cars you mention could possibly qualify as a “family car”. They all have fairly limited range and are best deployed as a commute vehicle that are used in a constant, predictable fashion. While I despise Top Gear UK for slagging electric vehicles and “demonstrating” how poor they are by using them in ways quite contrary to how they were marketed (equivalent to attempting to carve the Christmas ham with an electric toothbrush), folks like YOU are exactly the root of the problem, trying to promote electric vehicles for all usages from the start, whether appropriate or not. The Leaf et al. simply are not flexible enough to serve as the only car in a one-car household.

      As for the Model S, the only versions being delivered right now are models in the $90K -100K range with an EPA range of 265 miles (300 miles claimed by Tesla). Even when full production occurs, the cheapest Model S will be $50K (160 miles claimed by Tesla; if we extrapolate from the ones delivered, will get you 140). How many families do you know who can afford vehicles at those prices? All the tax credits available don’t make much of a dent in that pricing.

  11. Chris Carleton

    Thanks for this article. It’s one of the best pieces on batteries I’ve seen in a very long time. A balanced, scientifically based view without the usual arm-waving in either direction re: doom and gloom or heralding the second coming. It’s in the science … and we’ll keep moving in the right direction. There remain many challenges; fortunately, there’s a lot brilliant, innovative people dedicated to overcoming them.

  12. Einav Ibeiro

    Aluminum-water chemistry has specific-energy and energy-density comparable to gasoline. Recently there has been some proposals on how to utilize that chemistry for hybrid vehicles. See
    I am not sure about efficiency and cost though.