MacBook Air battery woes show the downside of the quest for denser batteries

MacBook Air battery woes show the downside of the quest for denser batteries

Over the past few years, manufacturers like Apple and Samsung have regularly delivered higher and higher battery densities, often with minimal increase to product weight or size. Apple was forced to make the iPad 3 slightly heavier and thicker when it introduced its Retina display, but it managed to increase total battery power by 70% while only bumping internal volume up by a fraction of that amount. As we discussed earlier this week, Samsung has dramatically increased the energy density of the Galaxy S4′s battery (as has Apple with the iPhone 5), again, without a corresponding leap in volume or weight.

But there are signs that this increased energy density may come at a cost. Over at his blog, IfWeAssume, developer Jared Davenport has documented his 2012 MacBook Air’s rapidly declining battery life. For the past year, Davenport has taken battery lifespan readings on a minute-by-minute basis, giving him an extremely detailed picture of the laptop’s battery health and remaining capacity. The results indicate a battery with a rapidly-declining capability — but that’s not the most interesting thing. Graphed against his 2009 MacBook Pro, the rate of decline is stark.

Image courtesy of IfWeAssume

There’s actually some reason to think that Davenport may have a bad battery; aggregate data from the website Coconut-Flavour, which tracks battery life on Macs, implies his battery is failing rather more rapidly than it should. But, at the same time, there’s also reason to think that the battery work Apple and Samsung have done could easily have a downside. Everyone has heard the phrase “there’s no such thing as a free lunch,” but when it comes to batteries, the physics are particularly unforgiving and the mathematics (particularly when it comes to things like failure rates) get really, really ugly.

The data below is drawn from Coconut Battery Online’s database of battery life over time.

It’s not hard to see that there’s two entirely different curves at work. At 36 months, the Macbook Pro has lost ~10% of its maximum battery power. The MacBook Air has lost 10% at 12 months. This may partly reflect different use cases in aggregate, though Davenport notes that he used both systems in precisely the same way.

Apple is extremely tight-lipped about its battery technology, as are the other major vendors. We know that the company has filed for multiple battery patents and considered different methods of boosting performance, but we don’t know if those filings, which date to 2009, have actually been used in any shipping hardware. 

Image courtesy of MpowerUK

As such, we’re going to talk in general terms. Packing more energy into a battery is actually less an issue than you might think. The real problems start creeping in when you want a battery that can be recharged several hundred times without losing much of its original capacity, fits into a tight space with minimal ventilation, won’t suffer thermal runaway, releases no toxic gas, and can’t expand when hot. You also want a battery that’s relatively resistant to overcharging or complete discharge, though this can be mitigated somewhat by building sophisticated software and charging mechanisms into the underlying equipment.

What a little digging into battery tech shows, however, is that the tradeoffs are rampant. We’re used to thinking of battery life in terms of power consumption vs. run time. Designers have to grapple with run time, weight, volume, anode/cathode size, discharge rate, voltage, temperature, and the chemical composition of the battery itself. Discharge temperature deviations that are totally fine with one battery chemistry may create sub-optimal performance with another.

One of the negative consequences of the smartphone boom has a relentless push to iterate on yearly product cycles. The problem with this approach is that it requires very careful timing to ensure that 2014′s battery technology improvement that you started researching in 2010 is actually ready for deployment in 2014. If something goes wrong, your company is stuck with a substandard solution compared to its rival — or it has to push out a product that might not quite hit the parameters it wants.

If you’re facing the choice between a slightly higher failure rate or shipping with no improvement to key metrics on your yearly refresh cycles, you may choose the first and call the warranty replacements a cost of doing business.