Nissan Leaf SoC: Fact or Fiction?

Recently, my beloved Prius finally gave up the ghost (brake pump died, over $5000 and months of repairs), so I decided it was time to replace it with an electric car. Specifically, a 40kWh Nissan Leaf.

And I highly recommend it.

There’s a lot of reasons to like it, not least because its fun to humiliate gas car drivers at stoplights, but because it has a fancy app which is a virtual treasure trove of data. As a side note, Nissan, stop messing with Tobias Westergaard Kjeldsen - I know you want to keep a tight fist over everything, but at the same time, this guy’s work makes me love my car even more (even if you think it exposes the uglies). And if I’m going to pay a monthly fee to use your app (which, oh by the way, doesn’t work on my phone), I demand access to the numbers.

Anyway, this means I can collect graphs in Grafana of things like power usage per day, miles driven per day, and battery percentage on an hourly basis, which if you’re an avid reader of this blog you would already know is the first thing I did with this car. So I started a new job recently, with a 2.6 mile commute, and this last week I finally went a whole week straight without driving anywhere super far away so I can collect some good data on how commuting affects my battery levels.

So here’s my miles and power consumption per day:

All fine and good. This comes to a total of 34.02 miles driven, and 7.45 kWh consumed, for an average of 4.57 miles/kWh. Which sounds right, based on what the dash tells me after I finish driving.

At the same time, I measured the battery state-of-charge (SoC) percentage over the same time period. I started the week at 55% SoC, and ended it at 29% SoC.

With some simple math we can extrapolate that “100%” SoC is equivalent to… 28.65kWh and 131 miles. For a 40kWh battery. The EPA range for this car is up to 149 miles, so the mileage is about what I would expect (I’ve been driving with the A/C on, which takes about 10% of the range). But that energy number is shockingly low. A couple of possibilities I can think of: * The car is lying to me about how much power I’m consuming, * Nissan padded the battery percentages, so “0%” and “100%” are actually closer to (if I had to guess) “18%” and “90%”, or * I’m bad at data collection and/or math.

My suspicion is the second one, although I’m surprised they didn’t then market it as a 30kWh battery which they then overprovisioned on the backend. But, we have another data source we can use to measure this: the chargers tell me exactly how much power I consumed (weirdly, though, they don’t bill me based on power, but based on time spent charging). Last Saturday, I charged 24.857 kWh to go from 23% (estimated) to 85% charge.

…which comes out to a “100%” of almost exactly 40.1kWh. The start point is not exactly known, but in order to hit the “30kWh per 100%” metric above, the SoC would have had to start at 3%. And the SoC was definitely not at 3% when I started charging.

So therein lies the mystery. Does the Leaf lie by almost a third about its mileage? This wouldn’t be the first time I’ve seen a car lie about its mileage. It wouldn’t necessarily be false advertising - the car does still hit the EPA-rated range - but it would throw off a lot of my estimations. And, surely people would start to piece things together, even without direct access to the data. But ChargePoint seems to agree with this conclusion.

Or maybe the Leaf’s battery SoC is nonlinear in a way that’s screwing with my analysis? This would be a logical feature to add to the car, like how a gas car’s fuel gauge shows empty for several miles before it runs out.

Either way, more experimentation is needed to determine what’s really going on. Also, really, maybe I need to get myself that “Leaf Spy” app and get more direct metrics on the battery. Stay tuned.