EPA Range calculation confusion

[brulaz]

Apologize for the nerdy first question, but I see a lot of tech types here so ...

Looking at the 2019 Tesla Model 3 AWD Long Range, for example:

EPA says it has a 310 mile range with 29 kWh/100mi combined (hwy+city) efficiency.
But with a 75 kWh battery, I only calc a 259 mi range.

Similarly if I calculate ranges using the reported city or hwy efficiencies (MPGe, and using the EPA's 33.7 kWh/gal) I get city and hwy ranges substantially lower than 310 mi.

First thought was that I have the reported battery size wrong. But the same thing happens with any other electric vehicle I've looked at. Don't think I have ALL their reported battery sizes wrong.

Could it be that the manufacturers are under-reporting their actual battery capacities?

If so, could the EPA range estimate be used to back-calc a more realistic battery size? Which I could then use to calc a HWY-only range estimate using EPA's HWY efficiency, for example?

?? Confused ??

 

[Pushmi-Pullyu]

brulaz, Welcome to the EV world of fuzzy kWh numbers.

Two reasons why the math likely doesn't add up:

1. The EPA estimate of kWh/100mi is an estimate of power drawn from the wall, which includes charging losses, which in real-world use are generally reported as approximately 15%*. Contrariwise, the EPA range rating is based on actually using the energy stored in the battery pack to power the car, which doesn't involve charging losses. So those two figures will never match... or at least, that's my understanding.

*Even that number may vary from charging session to session. If my understanding is correct, as a battery pack approaches full charge, the charcing efficiency drops. So if you do a partial charge starting with a mostly "empty" battery, you may experience perhaps only a 10-12% charging loss.

2. No EV maker reports both full battery capacity and usable battery capacity for their EVs. All EVs reserve some amount of capacity to prevent the batteries from being either fully charged to 100% or fully discharged to 0%; either is very bad for battery life. Tesla has in the past reported a nominal full capacity in its designations such as Model S75 or Model X100, but those numbers are rounded off to the nearest 5 kWh, so those aren't that accurate either.

P.S. -- You'll find a lot of argument over the "real" capacity, the full capacity, of the LR Model 3 battery pack. Is it 75 kWh? 78 kWh? Even 80.5 kWh? Here's one article on the subject, and please note I make no claims it's the last word:

From Edmunds.com: "2017 Tesla Model 3 Has 80.5-kWh Battery, According to EPA Filing"

* * * * *

Auto makers treat the amount of exact capacity their EVs hold in reserve as proprietary info, so they aren't going to tell you how much that is. I've seen a detailed claim that Tesla cars reserve between 4% and 8% capacity, along with a claim that different sizes of battery capacity cars reserved different percentages. But frankly I seriously doubt the 4% figure as it's awfully low. And in fact, I seriously doubt Tesla reserves more capacity in one battery size than another, for the same model. Why would they? It makes no sense, at least not to me. (Exception: Some Tesla cars have been sold with an electronically limited battery capacity, so those cars have a considerably higher reserve capacity which is never used.)

Anyway... Sorry to say, you'll have to settle for estimates. You can of course measure your own car's energy usage and efficiency, but you'll need third-party apps and/ or meters to measure that. There are dedicated EV owners who have installed a meter on their EVSE (the external charger, usually mounted on the wall), to measure the actual amount of kWh it's drawing when it charges their car. And I think there is some sort of smartphone app you can get which will tell you just how much energy your Tesla car is using when driving. Sorry I can't point you to the name of the app, but altho I'm a Tesla fan, I don't own a Tesla car.

And never, ever apologize for asking questions! How else can we learn?

 

[DucRider]

2. No EV maker reports both full battery capacity and usable battery capacity for their EVs. All EVs reserve some amount of capacity to prevent the batteries from being either fully charged to 100% or fully discharged to 0%; either is very bad for battery life. Tesla has in the past reported a nominal full capacity in its designations such as Model S75 or Model X100, but those numbers are rounded off to the nearest 5 kWh, so those aren't that accurate either.

* * * * *

Auto makers treat the amount of exact capacity their EVs hold in reserve as proprietary info, so they aren't going to tell you how much that is. I've seen a detailed claim that Tesla cars reserve between 4% and 8% capacity, along with a claim that different sizes of battery capacity cars reserved different percentages. But frankly I seriously doubt the 4% figure as it's awfully low. And in fact, I seriously doubt Tesla reserves more capacity in one battery size than another, for the same model. Why would they? It makes no sense, at least not to me. (Exception: Some Tesla cars have been sold with an electronically limited battery capacity, so those cars have a considerably higher reserve capacity which is never used.)

Anyway... Sorry to say, you'll have to settle for estimates. You can of course measure your own car's energy usage and efficiency, but you'll need third-party apps and/ or meters to measure that. There are dedicated EV owners who have installed a meter on their EVSE (the external charger, usually mounted on the wall), to measure the actual amount of kWh it's drawing when it charges their car. And I think there is some sort of smartphone app you can get which will tell you just how much energy your Tesla car is using when driving. Sorry I can't point you to the name of the app, but altho I'm a Tesla fan, I don't own a Tesla car.

And never, ever apologize for asking questions! How else can we learn?

​

BMW actual DOES report actual and usable on it's i3 batteries.
60Ah 21.6kWh actual/18.8 usable
94Ah 33.2/27.2
120Ah 42.2/37.9
120Ah specs here:
https://www.press.bmwgroup.com/global/article/attachment/T0284828EN/415571

And I believe battery health (capacity) is available thru a menu in the touchscreen.

I'm a little confused as to why manufactures are not required to disclose both the actual and usable capacities, as well as make the degradation status/information readily available when it is covered under warranty (some manufacturers - Tesla is one - specifically exclude degradation from their battery warranty)

 

[Pushmi-Pullyu]

BMW actual DOES report actual and usable on it's i3 batteries.
60Ah 21.6kWh actual/18.8 usable

Thanks! Good to know there is at least one EV maker which is transparent with this info.

I'm a little confused as to why manufactures are not required to disclose both the actual and usable capacities...

It would be nice if they did, but I certainly am not in favor of heavy-handed government regulations which would force them to disclose that proprietary info. Companies should be allowed to keep their trade secrets so long as it doesn't involve something which endangers the public.

I note that it's been claimed, apparently with some justification, that the "cubic inches" reported as the inside volume of some gasmobile engines is likewise subject to some amount of rounding off, rounding up, or outright exaggeration by some gasmobile makers. If they aren't required by law to accurately state the number of cubic inches in their ICEngines, then why should EV makers be required to disclose the capacity of their battery packs?

Just my opinion, of course. YMMV.

​

 

[DucRider]

When the manufacturer elects to provide a written warranty for xx% of original capacity, the consumer has a right to know:
a) the original capacity
b) present capacity

If only the manufacturer has access to that info, the consumer has no recourse or way to file a warranty claim.

Nissan skirts this by using the term "bars of capacity" (which a readily displayed) in it's warranty. They are, however, free to change what a bar represents, and have done so in the past to "restore" capacity to degraded batteries (I believe they claimed to be correcting a software/firmware error in some instances)

 

[Pushmi-Pullyu]

Nissan skirts this by using the term "bars of capacity" (which a readily displayed) in it's warranty. They are, however, free to change what a bar represents, and have done so in the past to "restore" capacity to degraded batteries (I believe they claimed to be correcting a software/firmware error in some instances)

I applaud Nissan for, belatedly, including an actual capacity loss in its battery pack warranty. However, let's not forget that this is a fairly recent change. Originally, the Leaf's battery pack warranty included replacement only for loss of power, not capacity/range. Of course, if enough capacity is lost, then power will also be reduced somewhat, so it's not like you could have a severe capacity loss without power loss. I'm not sure there is an exact ratio between loss of capacity and loss of power, but certainly the two are connected.

* * * * *

Personally, I see nothing whatsoever wrong with Tesla abandoning all specifications for battery pack capacity, either full or usable. It certainly simplifies things to talk about only miles of range, and how fast the car can be charged (in terms of miles of range added per minute or hour). In fact I applaud that simplification, as it makes it easier for the average car buyer to understand how an EV operates, and thus removes one obstacle to people making the transition from gasmobile to plug-in EV.

Sure, those of us interested in the engineering aspects are gonna want to know the actual kWh, the energy usage in terms of kWh per mile, etc. etc. But Joe and Jill average don't need to know the exact size (in gallons) their gasmobile's gas tank holds, and neither do they need to have a detailed understanding of the function of an internal combustion engine, merely to drive a gasmobile car.

Early motorcar enthusiasts did need to know how their (much, much simpler) motorcars functioned on a mechanical and electrical basis, because those early motorcars were prone to frequent breakdowns, and the driver needed to know how to diagnose and repair problems in the field. One of the ways that motorcars have improved over the decades is in becoming much more reliable. Similarly, as EVs improve, there should be less and less need for the average driver to know about such things as the kW rating of the motors, and kWh needed for charging or for range... or even the need to understand the difference between kW and kWh.

​

 

[DucRider]

Not knocking Tesla's approach (but the drop in estimated available range in cold weather does cause concerns about battery failure in new users of all brands if you watch the forums or attend EV events).

But, when a company elects to prove a warranty based on a % of capacity, they have an obligation to make that information readily available to the consumer. A manufacturer could simply say "it's still within specs" and the consumer would have no way to verify or refute that claim, even if anecdotal evidence pointed to the contrary.

In example, in a Bolt forum a user is claiming an estimated range of 76 miles when fully charged (in sub-freezing temps), but reports a 3.9 mi/kWh efficiency reading. Those 2 numbers point to ~20kWh capacity, but would it hold up for a warranty claim? (Bolt warranty is 60% of the original 60kWh). We don't know how much current temps are mixed into the range estimate, what weight the algorithm gives past driving efficiency, etc. Any range estimate must utilize at least some measure of battery capacity in order to make that calculation. Why not make that number available?

If the manufacturer elects to warrant a % and/or kWh capacity, there needs to be a way for the consumer to measure it.

Tesla elects not to use kWh or HP ratings on their cars (but they do bill by the kWh @ Superchargers where they can), and does not warrant any particular battery capacity or range - only against "failure" (if you can drive 50 miles, has it "failed"?). That's the way Nissan used to do it until llawsuits forced them away from that model. Tesla hasn't had significant battery issues, but if they did they would likely have to define "failure" of the battery in a non-subjective way (just as Nissan was forced to).

 

[Pushmi-Pullyu]

If the manufacturer elects to warrant a % and/or kWh capacity, there needs to be a way for the consumer to measure it.

I guess my comments above suggest that I'm against it. But I'm not at all opposed to it; I just think the industry is moving away from specifying kWh capacity, that's all.

Whether or not the guvmint should be able to mandate an EV maker revealing both the full capacity and the usable capacity of a battery pack... Well, that's a real can of worms. Even aside from the political issues, who is to say just what the "real" capacity (either full or usable) of a battery pack is? Ultimately, it's up to the battery cell maker to specify a nominal full capacity for a cell. But note that capacity varies somewhat by temperature! So how can anyone point to a specific number and say "That's the actual capacity of the cell"? The actual measured capacity depends on the testing conditions.

I suppose the EPA could specify conditions for a standardized method of testing battery capacity, in the same way they specify conditions for a standardized method of testing range. But how would that benefit the average car buyer? The important questions are range, the amount of energy required to charge your car (which includes charging losses, so it's not actually a measure of the battery capacity), and energy efficiency in terms of miles/km per kWh (or kWh per mile/km).

I take your point that if the auto maker guarantees a specific percentage of battery capacity remaining after 8 or 10 years, it's going to be hard to measure that without knowing the exact capacity is when the car is new... altho one could argue that the auto maker should only have to provide a number for the usable capacity, not the full capacity, because it's the usable capacity which limits how much range the car has.

I find it interesting that GM gave so much reserve capacity to the Volt 1.0's battery pack that nobody has reported any loss of capacity due to aging -- or so GM claims, anyway -- but that (according to a few reports) there are now some reports of a loss of power, which presumably is due to aging.

Tesla hasn't had significant battery issues, but if they did they would likely have to define "failure" of the battery in a non-subjective way (just as Nissan was forced to).

I'll argue that Nissan was "forced" to do so because of their inferior battery tech. Contrariwise, Tesla has done a great job of standing behind their warrenties, going out of their way to look for potential problems and replacing (for example) drive units before potential problems turned into real ones. I haven't seen any claims of Tesla refusing to replace a battery pack that has gone bad prematurely. Not even on the Tesla Motors Club forum, where even minor problems receive a lot of attention and discussion.

At some point, I think we need to let market forces do their work. An auto maker which gets a rep for not standing solidly behind its warranties is going to get a bad image in the market, and is going to lose sales. It will either correct the problem, go out of business, or see its market shrink down to a niche.

That's one reason why I'm all for capitalism and a competitive market. (But not unrestrained capitalism nor a completely "free" market. Totally free markets lead to monopolies, price fixing, and other non-competitive practices.)

​

 

[bwilson4web]

BMW actual DOES report actual and usable on it's i3 batteries.
60Ah 21.6kWh actual/18.8 usable

So I have a BMW i3-REx with these EPA specs:

• 29 kWh/100 mi
• 72 mi electric range
• (72/100) * 29 kWh = 20.88 kWh
  • (71.5/100) * 28.5 kWh ~= 20.38 kWh
  • (72.5/100) * 29.5 kWh ~= 21.39 kWh

So a sensitivity analysis suggests this approach is closer to the battery maximum, not usable capacity.

Thanks!
Bob Wilson

 

[brulaz]

Thanks for all the replies.

For simplicity I'm going to assume the advertised battery size is the "usable" size, which is pretty much all I need.

...
The EPA estimate of kWh/100mi is an estimate of power drawn from the wall, which includes charging losses, which in real-world use are generally reported as approximately 15%*. Contrariwise, the EPA range rating is based on actually using the energy stored in the battery pack to power the car, which doesn't involve charging losses. So those two figures will never match... or at least, that's my understanding.

*Even that number may vary from charging session to session. If my understanding is correct, as a battery pack approaches full charge, the charcing efficiency drops. So if you do a partial charge starting with a mostly "empty" battery, you may experience perhaps only a 10-12% charging loss.
...

Many thanks.

When I correct the EPA hwy and city MPGe by that 15% charging loss, I get more reasonable hwy and city range estimates that bracket the EPA's reported combined range. And I'm pretty sure I can fine tune a MPGe correction factor for each vehicle.

It's the hwy range estimate that I want. We don't have an electric vehicle yet, but if/when we get one, it has to be able to make a non-charge-stop, all hwy drive to the big city. Once there, city driving should be no problem.

We have other, prolly more important, range concerns as well, especially winter driving and loss of range from age. But those are for another day.

 

[Pushmi-Pullyu]

It's the hwy range estimate that I want. We don't have an electric vehicle yet, but if/when we get one, it has to be able to make a non-charge-stop, all hwy drive to the big city.

One of my pet peeves about the EPA is that they give only one number for an EV's "highway range". That's highly variable by speed. I wish the EPA would use a chart for estimated ranges at speeds of (for example) 35, 45, 55, 65, 75, and perhaps even 85 MPH.

As I understand it, the EPA's highway range tests average about 55 MPH. So if you're driving between cities at a speed of ~70-75, as most American drivers do, then -- fair warning! -- you may be quite disappointed at the range of your EV.

My armchair engineer rule of thumb advice for prospective EV buyers is to get an EV with 40% more EPA range than you think you'll need for your daily drive. That should cover most loss of range from highway driving at higher speeds, as well as the range loss from driving at city speeds in bitterly cold conditions.

We have other, prolly more important, range concerns as well, especially winter driving and loss of range from age. But those are for another day.

In reports from the recent "polar vortex" weather conditions, a lot of Model 3 owners experiencing their first winter expressed shock at a range loss of ~40%, or even in some cases up to 50%. I think that is a pretty extreme case, though; most comments I've seen indicate a loss of no more than 30% even in bitterly cold conditions. But there are quite a lot of variables. If you make multiple stops and leave your car sitting outside for hours during a day's drive, then you may well experience a 40-50% loss of range.

But it really depends on your driving pattern. Pre-condition the car so the battery pack is warm when you start, and don't leave it sitting outside without being plugged in for hours, and you will likely experience only a 20-30% drop in range, even on very cold days, and even if you don't forgo use of the cabin heater as some of the more dedicated EV drivers do. Personally I'm not that dedicated; I'd use the heater when appropriate!

Regarding range loss due to aging... for Tesla cars, that doesn't seem to be much of a concern. Unless you're planning on keeping your car more than 7 years, the range loss shouldn't be more than 10%, or 15% at the worst. That's not negligible, but it's far less important than range loss due to either high speed highway driving or driving on very cold winter days.

​

 

[bwilson4web]

As I understand it, the EPA's highway range tests average about 55 MPH. So if you're driving between cities at a speed of ~70-75, as most American drivers do, then -- fair warning! -- you may be quite disappointed at the range of your EV.

My understanding is the EPA adds an "engineering factor" that brings this close to ~65 mph. At least that has been my Prius and BMW i3-REx experience.

Bob Wilson

 

[brulaz]

...
As I understand it, the EPA's highway range tests average about 55 MPH. So if you're driving between cities at a speed of ~70-75, as most American drivers do, then -- fair warning! -- you may be quite disappointed at the range of your EV.
...
If you make multiple stops and leave your car sitting outside for hours during a day's drive, then you may well experience a 40-50% loss of range.
...
Regarding range loss due to aging... for Tesla cars, that doesn't seem to be much of a concern. Unless you're planning on keeping your car more than 7 years, the range loss shouldn't be more than 10%, or 15% at the worst. That's not negligible, but it's far less important than range loss due to either high speed highway driving or driving on very cold winter days.

All good points, and thanks.
So it looks like we need something with at least 300 mi EPA range which really limits the selection. Even when ignoring the affordability issue.
I was hoping the Bolt or Kona Electric would work, but now it seems that something like the Tesla 3 Long Range would be better.
Prolly as battery prices drop, others (VW, Ford?) will also come out with longer range vehicles.

 

[bwilson4web]

I am less worried about the range than the charging speed. Ultimately that determines how long the car has to sit at a charging station. That 'not rolling' time determines the block-to-block speed ... how long it takes to complete a trip.

The range only counts when measuring the distance to the next charger.

Bob Wilson

 

[Pushmi-Pullyu]

The range only counts when measuring the distance to the next charger.

And unless your overnight charging (slow charging at home/work) is limited to L1, charging speed rarely counts except when fast-charging on a long trip.

 

[Pushmi-Pullyu]

My understanding is the EPA adds an "engineering factor" that brings this close to ~65 mph. At least that has been my Prius and BMW i3-REx experience.

So what, based on your actual experience, should be the rule of thumb for a prospective buyer regarding the EPA range rating? I suggested the EPA range needs to be 40% above what you need for your daily drive. What would you recommend, Bob? All my info is "armchair learning"; yours is based on actual experience driving EVs, so I wonder what you would recommend.

And what's the climate like where you live? I see your profile says Alabama. How often, if ever, does it get bitterly cold (0° F or below) on winter days, or is the climate there more mild?

​

 

[bwilson4web]

So what, based on your actual experience, should be the rule of thumb for a prospective buyer regarding the EPA range rating? I suggested the EPA range needs to be 40% above what you need for your daily drive. What would you recommend, Bob? All my info is "armchair learning"; yours is based on actual experience driving EVs, so I wonder what you would recommend.

And what's the climate like where you live? I see your profile says Alabama. How often, if ever, does it get bitterly cold (0° F or below) on winter days, or is the climate there more mild?

I find the EPA range numbers are fairly close at 65 mph. It goes up with slower speeds and down, significantly at 70 mph and above. Regardless, I'm one of those owners who takes the EPA roll-down metrics combined with the powertrain efficiency and tank/battery to calculate my own mph vs range and mph vs hp charts. I calibrate the curves with benchmarks.

Alabama has mild winters with temperature seldom below 20F (-7C). Still, my cold weather benchmarks reveal that temperature becomes significant at and below 50F (10C.) It is measurable and becomes more so as the temperature decreases.

The reason I mention benchmarks is I tune my cars for optimum performance (i.e., highest efficiency.) I drive my cars to minimize operational cost and have had modest success. I reject the "hypermiling" practices even though I have benchmarks showing they can work ... too d*mn dangerous in traffic!

Bob Wilson

 

[brulaz]

I am less worried about the range than the charging speed. Ultimately that determines how long the car has to sit at a charging station. That 'not rolling' time determines the block-to-block speed ... how long it takes to complete a trip.

The range only counts when measuring the distance to the next charger.

Bob Wilson

You are right, and if there were a relatively fast charger close to a decent restaurant on our route to the Big City, that would work for us. And allow us to buy a more affordable vehicle with less range.

There actually is a Tesla Supercharger on the way but it wasn't really located near any restaurants we like. But now I just googled it and that area has grown up and looks a lot more interesting. I'll have to drop by and check it out. Also, I recently noticed a sign indicating the future location of a new non-Tesla charge station on our route that could work.

So these things are changing as well. And perhaps we can go electric sooner than I thought.

You folks are great helping me think this through.

 

[TheMagster]

Another gotcha that often isn't mentioned...you can't use the car's heater while it is Lvl 3 quick charging! This can make for a chilly half hour if you need to quick charge in the middle of the night during the winter and all the restaurants are closed. Best to avoid that scenario if you can. You can use the heater when Lvl 2 charging, but that takes hours so you are unlikely to be sitting in your car during that time.

At least that is true of my 2015 Leaf. Perhaps other EVs can do it, I'm not sure.

Sent from my Pixel XL using Tapatalk

 

[bwilson4web]

you can't use the car's heater while it is Lvl 3 quick charging!

My BMW i3-REx is happy to do this BUT the trick is to enable the heater when the battery is entering 'taper down.' I've seen a quick jump in the charging current.

Bob Wilson