First, a note for EV newcomers
The math below is just geeky fun. It is NOT an essential part of owning a plug-in vehicle; if you find it boring, ignore it. Driving electric really is simple. Here’s all you need to know to start using a plug-in without ever doing any range calculations or having to worry about hypermiling:
1. Buy a plug-in hybrid, like the Chevy Volt, Ford Fusion or C-Max, Fisker Karma, etc. They are still gas cars, so they drive just like a gas car on a trip. No thinking ever required.
- OR -
2. If you want to use even less gas, do less maintenance and save more money on a full battery-electric, look at the EPA range rating (not the higher marketing number. For example, Nissan often says the Leaf can go 100 miles. Tesla sometimes says the Model S can go over 300 miles. But the EPA ratings are 73 and 265). Now, subtract my magic-rule-of-thumb 1/3 to account for really bad weather or really high speeds (presumably you won’t do both at the same time). That leaves 49 miles for the Leaf, and 177 miles for the 85kWh Model S.
Those are numbers you can “count on” in your new car. Obviously you can do worse if you take it to the track or something, but as long as you aren’t trying to fail and really want to get somewhere, you will be able to go at least that far - and much farther once you get used to it. You can take the car on any trip of that length without worrying or doing math; so just make sure your next Model S charging stop is never more than 177 miles away. Math is only required if you want to go farther than that without stopping to charge.
After several (5-10) years the battery will lose some capacity, so we need to account for that too. Of course by then you’ll have lots more experience and will be comfortable taking the car on longer trips; but if you have a long commute and want to make sure the car you buy will ALWAYS work for you without ANY planning or charging, then take 70% of the number above. That puts the Leaf down to 34 miles, and the 85kWh Model S down to 124 miles. If you have to drive farther than that, don’t have another car for those long trips, don't want to rent, won’t ever be able to charge, and don’t want to buy a new battery, then you will be more interested in a plug-in hybrid.
Let the math begin
OK, now that we are only left with people that WANT to go through the math, here are some numbers for calculating how far you can get in your Model S. The more you understand the car, the farther (and more often) you can take it rather than your old gas car that you really don’t want to drive anymore.
Spoiler: you can skip all the math if you always plan charging stops within 2/3 of the EPA range of your car - unless you're racing up a mountain, you will make it (although it's always wise to plan an extra 20 mile buffer or so, just in case the charger is unusable!). For the 85kWh car, that's 177 miles. For 60kWh it's 139 miles, and for 40kWh it's 95.
The math is not complicated - it's just adding percentages together to see how it compares to rated range. At 100%, you should by definition expect the rated range and a 200-mile trip should be easy, leaving you with 65 miles left over. But if you add in, say extra speed and cold weather to get 133% of rated miles, that means your 200-mile trip will take 266 miles of rated miles - and you will likely not make it. Plan a charging stop.
When planning a trip in advance, I don't know what the weather is going to be and so I plan for the worst and never have charging points be more than 177 miles apart - the longest drive with no charging in my Death Valley trip was 175 miles, and even there I had a plan for a place to charge in the middle if things went bad. You can make assumptions and plan it closer than this, but in my experience that just adds stress. On the day of the trip, I know a lot more about conditions and so I can make adjustments - sometimes, for example, I take off from a charging stop before charging all the way because I know the buffer is much larger than it needs to be. Or sometimes I drive faster just because I know I have the extra range.
Range number origins
The Model S 85kWh is rated by the EPA to go 265 miles per charge. That rating comes from a complex test that includes city driving, speeding up and slowing down, etc. But when calculating range, most people are concerned about steady driving on the freeway. From Tesla’s numbers, 265 is roughly what you’d expect going 60mph on dry level ground with 19” wheels and a small amount of HVAC use (outside temps of, say, 55 or 90). 300 miles is what you’d expect going 55mph on dry level ground at 70 degrees with no HVAC.
The EPA number is a good number to use to compare vehicles. But it’s not a good number to use to tell how far you’re going to go on any specific trip. That’s why I wish EV battery indicators simply gave the % of the battery left, rather than a number of miles which is just the battery % multiplied by some factor that they don’t explain clearly and may not apply. It is true that new owners will find it confusing; but after a few weeks of using the car, it is much better information. New 85kWh Model S owners could simply be told to figure they’ll usually get 2 miles or so per percent of battery; perhaps up to 3 miles if they drive carefully in ideal conditions.
All of the numbers I'm going to use below assume you are taking a long trip. When you take short trips, initial HVAC load may be a much higher percentage. However, that is often balanced out by much lower speeds as well as the far greater unlikelihood of putting on more than 177 miles in the day. Short trips in a long-range car generally don't need calculations (people with short-range BEVs do these calculations more often; but it's not a big deal for them once they are used to the car because with short-range cars you typically make the same trips over and over, so you soon "know" if you can make it or not and can skip the calculations). Note that around-town consumption is often lower in BEVs because of the much lower speeds; but may well be higher in the Model S because of the much greater than average weight that needs to be accelerated from every stop, plus the awful temptation to take advantage of all the available acceleration.
Tesla’s range calculator is attractive and easy to use. It seems to be pretty accurate. It says that if you go 65mph on level dry ground with HVAC on at 32 degrees, you will get 218 miles. That’s using 118% of rated miles to go 5mph faster and use more HVAC. But what if you go even faster than 65? What if it’s colder than 32 degrees? What if there is rain or snow on the ground? What about elevation changes? I asked Tesla to add these items to their calculator long ago, but no word on when or if they will.
HERE is another calculator that takes elevation in to account, and allows a wider range of speeds to be entered. But it doesn’t account for weather or road conditions at all.
Somewhere on these forums I saw mention of yet another calculator that sounds interesting, but it is Mac-only and I don’t have a Mac so I have not been able to try it. (EFusco found it for me: it's HERE).
Update Mar 2013: a Model S owner (Cliff Hannel on these forums) has been building his own version HERE. Forum thread HERE. It is not finished yet, but it appears to consider all of the relevant factors (speed, temperature, elevation, etc) except road conditions. This would be a great web page to have saved in your car's web browser.
Tesla's consumption documentation and the calculators were a great place to start. But I had to extrapolate some of the numbers, and add in some real-world EV experience for things they did not address. In addition to a Prius I turned in to a PHEV in 2008, and a Toyota RAV4-EV I bought in 2009, and a Tesla Roadster I bought in 2009, my wife and I have been keeping track of numbers on our Model S that we received in September. We now have 7,000 miles on the car, and recently returned from a 3,000 mile trip.
Do not take these numbers as gospel! They seem to work pretty well for me, but I have done a lot of strange and varied things to get them. Several of them are really just guesses. I look forward to hearing from people that have thoughts on refining the numbers. Try 'em out and see if they work for you.
Numbers from our trip are below, but first here are some factors that I consider when estimating range.
This is the largest factor that you have control of. (Acceleration could be a far greater factor if you did it a lot; but it is typically of very short duration on a road trip). If you are ever in trouble, slow down. But if you plan, hopefully you will never be in trouble as your next charging point will always be easily in range.
I sometimes slow down to be more efficient; especially if my next charging stop is a slow one - going slower can reduce your wait at the next charging station. But I always plan my charging stops to be close enough that I never have to slow down to make sure I will make it. Worries about "having to drive a BEV slow" and "sweating bullets about making the next charging point" are overblown - don't push the limits, and it will never be a problem.
|60||100% of rated range (at 70 degrees with miniscule HVAC)|
At full blast, HVAC pulls about 7.5kW. What % hit that is depends on how fast you are going. And of course you will not usually have it on full blast. AC takes less than heat (at 90 degrees Tesla says the range is the same as 70; Model S batteries must like it a little warmer than Roadster batteries). But at 60 – 75mph freeway speeds, 30% would probably be the upper end. It appears about 1/3 of this hit happens even with HVAC off, so that’s the portion to manage the batteries. Because the HVAC draw is time-based rather than mile-based, these percentages should be slightly higher for lower speeds and slightly lower for higher speeds, but I haven’t taken the time to do the adjustments here.
If you ever get in trouble (again, plan for the worst case and you shouldn't) and the weather is extreme, you can turn the HVAC off and that can save quite a bit. It can be very unpleasant however, especially if you have a spouse grumbling about it the whole time. Also note that you only save 2/3 of the numbers below by doing that, as the car will keep conditioning the batteries.
A couple of times (when my wife is not with me) I have turned down the heat to be more efficient; especially if my next charging stop is a slow one - using less heat can reduce your wait at the next charging station. But I always plan my charging stops to be close enough that I never have to turn off the heater just to make sure I will make it. Worries about "being unable to use heat" and "sweating bullets about making the next charging point" are overblown.
|15 degrees F||+25%|
|70||100% of rated range (HVAC on, but barely in use)|
Update: Fleetcarma.com just released some data on how temperature affects the Leaf's range. Of course it's a different chemistry with no thermal management on the pack, so the numbers will be different, but it is still interesting. It too gets near-max range in 70 - 90 degree weather. 32 degrees seems to take off about 20%, and 15 degrees takes off about 30%. Those are bigger hits than on the Model S, but this was measured on the older Leaf that uses a resistive heater; the 2013 will have a heat pump. At -13 it took off about 40% (the cabin heater, which is not terribly strong, probably maxed out at around 15-20 degrees; but the pack heater kicks in at 14 degrees so greater draw continues). 100 degrees only seems to take off a percentage or two, but you can see the curve is getting sharper around there.
Other EVs (Roadsters and RAV4-EV's) use 7 miles up and 4 miles down; but the Model S is heavier. Subtract 10 miles for every 1000’ of elevation gain. Add 6 miles for every 1000’ of elevation loss. There’s probably no getting around the numbers for elevation gain, but there could easily be variation in the numbers for elevation loss – it depends on how fast you are going, how gradual the loss is, how much regen and wind resistance affect your speed, etc. Gain and loss numbers could both be larger if you carry more weight in your car.
Note that these are the number of miles to add or subtract, so you have to divide by the length of the trip to get a percentage. Ideally you should count ALL of the ups and downs on the trips; although I generally just look at the net elevation change, and any major passes along the way.
Stuff on the road (could be sand, rocks, garbage, varying pavement quality, or what have you; but most often rain and snow) obviously slows the car down – just FEEL how your car slows down when you’re coasting on the freeway and hit a puddle. I know from experience that stuff on the road makes a difference, but these numbers are really just guesses. They do seem to roughly fit the data, though.
|Clear and dry||100% of rated range|
This is very speculative*. But a few people that have driven their cars to low levels in dropping temperatures have gotten warnings that the number of miles remaining may suddenly drop. And many people have reported that if you leave the car sit overnight without being plugged in, a lot of miles (say, 10 - 30) may disappear. A couple of people have further noticed that if you get in that cold car that has lost miles and drive it, some of the miles seem to “come back” - they use fewer rated miles than expected to get to their destination.
It is possible that to protect the batteries, in colder weather the car may hide a certain percentage of the battery’s remaining energy, or of its total energy capacity. (It is also possible that it lists a lower number of miles assuming lower efficiency or more pack warming; but I think that less likely and doesn’t fit anecdotal reports quite as well). As the car warms up, some of the hidden energy may reappear. But how much, and at what temperatures? I don’t have nearly enough data to say with any certainty; this only happened to me a couple of times on this trip. But a 49-mile trip after a cold soak ended up doing about 4 miles better than expected as temps rose from 29 to 33 (that trip started at 173 rated miles, after losing 10-15 from the night before). A 70-mile trip after a cold soak did about 18 miles better than expected as temps rose from 30 to 42 (starting at 126 rated miles after losing 24 from the night before).
Fortunately it looks like the recovery roughly counteracts any phantom losses during the night; if so this may not be a measurable factor (except when you do your math based on the lower cold-morning numbers; but in that case it improves your numbers, so no harm done).
*UPDATE: This is a little less speculative now. Consumer Reports ran in to this (losing 30 miles overnight), and Tesla confirmed that the car "readjusts" range in freezing weather, and that it will sneak back as the car warms up. They didn't give exact amounts or say at what temps; but for simplicity's sake (and because it matches my data and what I've heard from others) I will guess the car really only loses 4 miles overnight in cold weather. If you check in the morning and it's below freezing, you will appear to have lost an additional 10 - 20 miles perhaps depending on the temperature delta. Those miles will, I assume, all come back if you get well over freezing; but you might only get, say, half of them back if you're still close to freezing. I'm not really confident in those numbers, but they fit the data we have so far.
Wind can affect you just as much as speed; it’s wind resistance that causes the drag. However it can help as much as hinder, is difficult to measure from inside the car, usually is not strong enough to be a large factor, and even when it is strong it typically doesn’t last the whole trip. So I haven’t accounted for wind here; but it’s entirely possible that it is responsible for a large bit of the differences between my expected and actual energy use. Wind has never caused me to run out of power in an EV, but the only time I ever ran out of gas in an ICE was because of a serious headwind in South Dakota that affected my mpg much more than I expected (I think I normally got 30-35mpg, but got just over 20mpg on that trip and fell short of the city I was planning to buy gas in).
Random other factors that can affect performance.
An early version of the calculator gave a much smaller loss (roughly 2.5%) due to the 21” wheels. I am not sure why the number is so much larger now.
Headlights and other 12V loads are minuscule. I don’t bother to calculate them at all.
|21" wheels instead of 19" wheels||+6%|
|Attaching pizza pans to your wheels||-6.5%|
|Rolling down the windows||+4%|
|Turning on headlights||+0.25%|
Applying it all to a real trip
Here are numbers from our 3,000 mile trip from Seattle to Death Valley in mid-Jan 2013. The first important number is 0; that's how much I spent on fuel for this trip. All of the chargers are free.
Our car in Death Valley:
This was not my first EV road trip; I've taken the Roadster on several long trips. My first trip, almost 3 years ago, was from Seattle to San Diego at a time when there were NO charging stations of any type North of San Francisco, and only a few L2 stations South of there. I was really excited to take the Model S as it is much larger, quieter, smoother, can use Superchargers when available, allows HVAC while charging, etc. Yet there is still room for improvement; for example, it still doesn't notify you when charging stops. That is a serious pain when you are charging at an L2 station for hours as you have to keep checking it to make sure a breaker didn't trip, or risk a several-hour delay to your journey (that has happened to me more than once, and is why I added OVMS to my Roadster).
Before the numbers, a mess of caveats:
- I did not account for wind
- I did not adjust HVAC consumption for speed
- We weren't exceptionally careful on writing down all the numbers exactly when we left and arrived
- We did not account for occasional heavy acceleration
- We did not even try to get precise measurements of water on the road, exactly how much time was spent at varying temperatures and speeds, etc
- I counted net elevation changes and major passes, but did not figure other hills
- I just added the percentages of the various factors to make the math easy. Their interaction is usually more complex; but my method probably overstates the hit, and so is safer.
- The car's Wh/mi readings did not seem to correlate well with the number of rated miles used. I find this surprising, and may start another thread to discuss that. For now I have those numbers grayed in the table so we can ignore them and just focus on the rated mile numbers, which are easier for most people to understand anyway.
No matter how much time you spend calculating, any number of things can still go wrong. You could hit very high winds, you could do the math wrong, the charging station could be blocked or broken...always leave as much buffer as you can. That's why Superchargers are 120 miles apart instead of 250 miles. You should do that with your charging stops too. It's less stressful, more pleasant, easier on the battery, and gives you options when bad things happen. If your charging stops are close enough together, you never have to do any math!
|Trip leg||Actual Distance||Temperature||Elevation change||Road condition||Speed||% of rated miles expected to use||% of rated miles actually used||Rated miles used||Wh/mi|
|Redmond WA 50’ to Centralia WA 200’||93 miles||37 - 43 F|
|63 - 70|
|133% of rated||139% of rated||129 miles|
|Centralia WA 200’ to Tigard OR 200’||100||40|
|Tigard OR 200’ to Eugene OR 500’||106||39 - 41|
|Eugene OR 500’ to Canyonville OR 800’||91||32 - 36|
|60 - 66|
|Canyonville 800’ OR to Yreka CA 2800’||124||27 - 34|
|40 - 65||141%||145%||180|
|Yreka CA 2800’ to Orland CA 300’||155||27 - 54|
|Mostly dry||55 - 70|
|Orland CA 300’ to Davis CA 300’||94||39 - 48|
|Davis CA 300’ to Palo Alto CA 150’||109||48|
|150' loss||dry||65 with traffic|
|Palo Alto CA 150’ to Gilroy CA 300’||49||29 - 33|
-8% cold soak
|119%||118%||58 (really 62 after cold soak)|
|Gilroy CA 300’ to Coalinga CA 450’||113||38 - 48|
|dry||55 - 74|
|Coalinga CA 450’ to Lebec CA 1100’||116||38 - 52|
|Lebec CA 1100’ to Barstow CA 2400’||152||37 - 52|
|dry||55 - 70|
|Barstow CA 2400’ to Furnace Creek CA -200’||175||25 - 47|
|dry||55 - 70|
|120% (no net elevation effect!)||126%||220|
|Furnace Creek CA -200’ to Barstow CA 2400’||175||52 - 64; little HVAC|
|dry||55 - 70|
|Barstow CA 2400’ to Mojave CA 2800’||72 (with detour)||49 - 59; HVAC off|
|dry||55 - 70 (more 70)|
|Mojave CA 2800’ to Lebec CA 1100’||70||30 - 42|
-29% cold soak
|dry||55 - 65||72%||74%||52 (really 72 after cold-soak)|
|Lebec CA 1100’ to Coalinga CA 450’||116||44 - 59|
|dry||70 - 80|
|Coalinga CA 450’ to Gilroy CA 300’||112||60 - 68|
|dry||55 - 85|
|Gilroy CA 300’ to Davis CA 300’||131||55 - 68|
|none||dry||55 - 75|
|Davis CA 300’ to Orland CA 300’||94||35 - 41|
|Orland CA 300’ to Yreka CA 2800’||155||45 - 70; little HVAC|
|dry||65 - 70|
|Yreka CA 2800’ to Canyonville OR 800’||124||32 - 45|
|dry||60 - 70|
|Canyonville OR 800’ to Eugene OR 500’||91||28 - 30|
|Eugene OR 500’ to Tigard OR 200’||106||29 - 37|
|dry||60 - 68|
|Tigard OR 200’ to Centralia WA 200’||100||36 - 41|
|none||dry||55 but mostly 74|
|Centralia WA 200’ to Redmond WA 50’||93||34 - 36|
|150' loss||dry||64 - 74|