Hacker News Reader

> Taken one step further, since Tesla is constantly transmitting telemetry to home base, Tesla HQ likely knows the average battery consumption of all other drivers who have ever driven that route. This would allow non-topographical variables (traffic lights, potholes, road conditions) to be factored into the range calculation as well.

This is super powerful. This in combination with a service like Waze in cars would produce some of the most accurate traffic and estimation data for travelling. I'd be interested to see Tesla incoroporating all these factors into their in-car estimation for ETA.

Tesla's already done things that used data collected in the past. For example, a year after I bought mine, they started labeling all of the chargers you've ever charged at on the map... all the way back to when your car was delivered. So I've got a charge icon sitting on the Tesla factory in Fremont, where they had it plugged in when I took delivery.

I suspect they have already started to do so.

Every morning I drive in the HOV lane of the 101. Waze and Google Maps give me traffic-based ETAs that are about 20 minutes longer than Tesla's. And every morning, Tesla wins. (And since I am a dork I continue to pit the 3 apps in this lopsided battle.)

The 20 minute delta is precisely the difference between the HOV and non-HOV lanes. My guess is that the Tesla traffic estimator is using telemetry from other cars on the road, which are also using the HOV lanes like me.

</anecdote>

This is purely speculative, but the technology certainly exists that would allow Telsa to do this.

All cars use more energy to climb hills. If you'd like to factor that into your Tesla planning, check out https://evtripplanner.com/ -- It shows the Vacaville/Truckee segment as taking 173 rated miles to cover 130 actual miles.

It's worth noting also that he's not talking about "driving in hilly terrain", the original comment is talking about driving almost exclusively uphill. Going to Tahoe you'd go from somewhere around 0ft to around 8500ft depending on the pass, and then down only to 6200ft.

Of course, this will impact all cars, though I find only a few MPG difference in our vehicles driving up to the ski resorts around Tahoe.

One curious thing I've noticed is that my relatively-high-Torque-and-HP car with a manual transmission gets almost the same mileage on flat ground versus going up and down mountain ranges, assuming there's no net elevation gain.

My girlfriend's fairly underpowered Subaru with a 4 speed automatic does significantly worse in the hills than it does on flat ground.

Also interestingly, the Tesla should get significantly better mileage in thinner air (on flat ground) where a gas-engined car has to deal with less air to burn which balances out the lower wind resistance.

Would also be interesting to see how the elevation change would affect a tesla at net 0. Say, drive from Sacramento to Truckee and back; I think the range should be plenty.

The website I referenced can do that computation for you. It's been accurate +/- 5 mi per supercharger segment for me. It estimates 216 rated miles to drive 201 actual miles, from Sacramento to Truckee to Sacramento.

> a gas-engined car has to deal with less air to burn which balances out the lower wind resistance.

Thats only true if you have an old carbureted car that requires turning a screw under the hood to adjust the air-fuel mixture. Modern cars measure the amount of O2 in the exhaust and adjust the mixture to keep the moles-of-O2 to moles-of-gasoline ratio roughly fixed.

The thinner air means you have to open the throttle wider to achieve the same manifold pressure, which in turn means that with the throttle wide-open, the engine will produce less power. But the thinner air should have no effect on the amount of gas you need to burn to achieve a particular amount of power.

If anything, you'd expect lower fuel consumption for a given amount of power due to the reduced pumping losses.

I'm pretty sure most cars don't get less than half the fuel economy in mountains than on flat roads, which is what the commenter claimed.

Most cars have a much lower MPGe than the Model S, so it's hard to directly compare the change in effort needed for steep uphills. Both Model S and "most cars" have to obey the same physics in order to lift the car to the top of the hill.

[Edit: you can think about it this way: A 89 MPGe Tesla Model S going up a steep hill at only 44 MPGe is still twice as efficient as a 20 MPG SUV driving on the flat.]

Agree: I get about the same (350Wh/mi) and hills make a huge difference.

That said, there is evtripplanner.com, which I found to be EXCELLENT at estimating range when I drove from Portland to Yosemite this summer. I summarized the estimate vs actuals in the Tesla forums:

http://www.teslamotors.com/forum/forums/made-it-portland-yos...

> Under normal conditions, I use 350 kWh/mi (kilowatt-hours per mile).

Say you get a good rate and pay $0.10 per kilowatt hour for electricity. That means if you charge your car at home, you're paying $35 per mile to drive. That doesn't seem right.

Possibly it's watt-hours? That'd be a per mile price of 3.5 cents, which seems about the right order of magnitude.

Wikipedia says one version of the model S has a battery pack with 85 kWh of juice. So, yes, at 350 kWh/mi, you'd get further on foot... :)

That sounds about right - the 85 kWh battery of the Tesla would then get you about 240 miles range.

Another interesting way of looking at this then, is that the amount of power required to power s single 100 Watt Light bulb for 3 1/2 hours could move an entire Tesla + person inside 1 mile. Wowsa.

Or: an 85kWh battery could power a 13 watt CFL for 272 days. Or power every light in my house for 18 days.

"could move an entire Tesla + person inside 1 mile"... on level ground.

If you've ever had to push a car manually in that situation, you'll notice that it's very easy to keep it moving once you get started, as you only have to overcome the rolling resistance. I've never pushed a Tesla before but I'd bet that it's far easier than a regular car, and a regular car isn't that hard to push already.

According to this page:

http://physics.ucsd.edu/do-the-math/2011/11/mpg-of-a-human/

A human walking takes between 100-200Wh/mi, so the Tesla isn't that far off.

> I've never pushed a Tesla before but I'd bet that it's far easier than a regular car

Based on what exactly? A Tesla Model S weighs at least 1000 lbs more than similarly sized cars.

Tesla's 6.1 software, announced Jan 8 2015, indeed predicts energy usage... although evtripplanner.com shows that you can do so pretty well using map data:

"To help you better anticipate your charging needs, Model S will now calculate the projected energy remaning at your destination using your predicted speed, the elevation, and other factors along your route. Trip Energy Prediction is only active while using Navigation."

I've noticed this, too. I have a plug-in prius, which only gives you about a 10 mile electric range, so the game ends up being "can I make this trip to X and back without having to burn any gas?" Knowing the terrain in your neighborhoods and knowing the paths that don't involve a lot of elevation change can really make a big difference in whether you "win" this challenge.

The range charts also show HVAC OFF which is highly unrealistic for most of the USA, most of the year.

I drove over the California grapevine in both directions last week and averaged 300 Wh/mi by sticking to 60 mph uphill and taking advantage of recharging while descending. Granted that climb is not as big the climb to Tahoe.

I'm finding the 2x difference in efficiency hard to understand. Gasoline cars get worse mileage in mountains, obviously, but it's nowhere near 2x. Is there some particular difference that would make the Tesla penalty so much worse? Perhaps the Tahoe area is also very cold, and Tesla's interior heating is a major battery killer?

On level ground, the majority of the energy goes into overcoming wind and drivetrain resistance, whereas going uphill requires a lot more energy just to "lift" the car (i.e. increasing its gravitational potential energy.) Since the Tesla is more efficient than a regular car on level ground, it makes sense that the difference between that and slopes will be higher too. On a regular car the efficiency is already low enough that the difference between hill-climbing and level ground operation is reduced.

The 500Wh/mi difference is essentially the energy needed to lift the car. Consider another hypothetical car, much less efficient, that took 2000Wh/mi on level ground; then it might take 2500Wh/mi going uphill, which is only 25% more.

Note that the Tesla will also recharge its batteries going downhill with regenerative braking, so it mostly balances out - as long as you don't completely flatten the battery going up.

Perhaps I'm not following, but doesn't that explanation imply that a Tesla's efficiency increase over the average car comes overwhelmingly from reductions in wind and drivetrain resistance? For the numbers to come out this way, that efficiency increase would have to be vastly larger than I would have guessed.

I would have guessed that most of the loss of efficiency (as in the fuel's heat of combustion divided by the displacement of the car's weight) of a gas car was heat loss from the internal combustion engine. But I know very little about cars.

There are other things.

For example: Teslas improve general efficiency considerably by (for example) using regenerative braking, which will be utilised in normal driving but not if you're travelling up a big hill on a freeway.

Furthermore, they're particularly heavy cars - a Ford Fusion (similar sized saloon) is about 1500kg, whilst the Tesla is 2100kg, presumably because of the batteries - the extra effort to lift up a hill is linear in the weight, so it'll be at least a 25% drop in the surplus cost over normal driving.

Overall, it's not really an efficiency thing because the costs of fuels vary so much. For pure heat-to-wheel efficiency, the most important one is that since Teslas are about 80% efficient at turning power into motion, and charge efficiency is about 97%, and a combined cycle power station is about 55% efficient at converting heat energy into power, then the total efficiency from heat to wheel is about 43%, whilst a modern petrol car might be expected to see more like 30%.

I don't know about the Tahoe area, but I get a lot less gas mileage when driving the hills of Los Angeles. It's not just the inclines, but even the flat-ish roads are very curvy, which involves a lot of braking and then acceleration.