Tesla confirms Model 3 will have less than 60kWh battery option

[JeffK]

With regards to EVs I don't think the average consumer pays attentions to MPGe. They just care about mi/charge.

We all drive differently and the new EPA estimates are conservative compared to other those of other organizations. Luckily the testing process is the same for other EVs. Comparing EV to EV is pretty straightforward.

 

[Red Sage]

This link is pretty good, and if I am reading correctly says that charging losses are not considered in range, but instead

1. (UDDS+HWFET) in energy/mile while driving * 5/7
2. Usable total battery capacity

The electricity consumption on the other hand does include a charging loss fudge factor of ~ 13% IIRC

Yeah. That's what I meant... The EPA uses two sets of numbers... One that is financially based... Another that is test results based... Then put them both on the Monroney Sticker using the same terminology based upon a cryptic 'MPGe' figure that is left up to the imagination to determine what it stands for. All the while, making every effort to make EVs of all types seem as inconvenient as possible.

 

[Jeff N]

This link is pretty good, and if I am reading correctly says that charging losses are not considered in range, but instead

1. (UDDS+HWFET) in energy/mile while driving * 5/7
2. Usable total battery capacity

The electricity consumption on the other hand does include a charging loss fudge factor of ~ 13% IIRC

My read of that article is that they drive the car on the dynamometer through multiple test cycle runs each time starting with a full battery and going at the end until the car poops out and falls below some threshold of propulsion. They then gather those results and run them through a formula and calculate the final EV range.

The energy consumption estimate comes from measuring the energy that it takes to fully charge the car between test cycle runs. The EPA does not literally use a "fudge factor" for charging loss nor do they determine charger efficiency or usable battery capacity.

 

[Red Sage]

My problem with the EPA is that they cannot decide what to be when they grow up (money Vs emissions,) and the attempt to appease ignorant Americans who have no idea what a kWh is but have no intention of buying an EV is nonsensical.

Yeah. Money spent for the sake of Consumer economy... Emissions expelled and how it affects the environment... Fuel consumption rates for vehicle fleets... And attempting to bind all these notions together in a unified MPGe Theorem of utter confusion. Having the data is nice, but how it is used and presented is of paramount importance.

 

[SageBrush]

My read of that article is that they drive the car on the dynamometer through multiple test cycle runs each time starting with a full battery and going at the end until the car poops out and falls below some threshold of propulsion. They then gather those results and run them through a formula and calculate the final EV range.

The energy consumption estimate comes from measuring the energy that it takes to fully charge the car between test cycle runs. The EPA does not literally use a "fudge factor" for charging loss nor do they determine charger efficiency or usable battery capacity.

From the Horse's mouth ...

A preview is available, or you can buy the full 26 page report for $74 ;-)

Addendum ... details. As usual, Jeff N was closest ...
EPA Test Procedures for Electric Vehicles and Plug-in Hybrids DRAFT Summary – Regulations take Precedence D. Good July 5, 2012

Note: FE Label format requirements have been revised in the 2013 Fuel Economy Labeling Rulemaking, (final rule 76 FR 39478, July 6, 2011)

1. Electric Vehicle Test Procedure - In general, EPA testing follows SAE Recommended Practice J1634 Issued 1993, "Electric Vehicle Energy Consumption and Range Test Procedure," which (as the title implies) is basically a dynamometer test procedure used to measure the energy consumption and driving range of an electric vehicle.

Electric Vehicle - City Test Procedure Summary - Following SAE J1634 Recommended Practice, the battery is fully charged, the vehicle is parked over night, and then the following day the vehicle driven over successive city cycles until the battery becomes discharged (and the vehicle can no longer follow the city driving cycle). After running the successive city cycles, the battery is recharged from a normal AC source and the energy consumption of the vehicle is determined (in kW-hr/mile or kW-hr/100 miles) by dividing the kilowatt-hours of energy to recharge the battery by the miles traveled by the vehicle. To calculate the energy consumption in units of mpge (miles/gallon equivalent) we use a conversion factor of 33.705 kilowatt-hours of electricity per gallon of gasoline (which is basically a measure of the energy in gasoline (in BTUs) converted to electricity). The city driving range is determined from the number of miles driven over the city cycle until the vehicle can no longer follow the driving cycle.

Electric Vehicle – Highway Test Procedure Summary - The same test SAE J1634 procedure outlined above, is used to determine the highway energy consumption and the highway driving range (except the vehicle is operated over successive highway cycles). Electric Vehicle - Adjustment Procedure used to Derive FE Label (Window Sticker) Estimates - EPA regulations require fuel economy, energy consumption, CO2 and driving range values listed on the FE Label (window sticker) to be adjusted to more accurately reflect the values that customers can expect to achieve in the real world. EPA currently allows fuel economy, energy consumption, CO2 values, and range values listed on the FE Label (window sticker) for electric vehicles to be adjusted using one of the following methods:1 • by multiplying city/highway fuel economy and range values by 0.7 and dividing city/highway energy consumption and CO2 values by 0.7; • using the vehicle specific 5-cycle method described in 40 CFR 600.210-12(a)(1); • using a method which is equivalent to the vehicle specific 5-cycle method described in 40 CFR 600.210-12(a)(1) (with prior EPA approval); • using adjustment factors which are based on in-use data (with prior EPA approval).

---
RedSage was right -- the range estimate includes charging losses.

 

[SageBrush]

bind all these notions together in a unified MPGe Theorem of utter confusion

LOL. I wish I could write half as well as you.

 

[Jeff N]

The EPA uses two sets of numbers... One that is financially based... Another that is test results based...

I'm not following what you mean. They do not in any sense use "financially based" sets of numbers. They measure the kWh that go into the car from a wall socket. That is an energy consumption number. It has nothing directly to do with pollution emissions or energy cost. Those are calculated separately based on actual electricity sources.

Then put them both on the Monroney Sticker using the same terminology based upon a cryptic 'MPGe' figure that is left up to the imagination to determine what it stands for.

I don't like MPGe either. I agree it is cryptic. They claim it was the best understood comparison metric among test groups of normals. Abnormal people like us would prefer Wh per mile.

All the while, making every effort to make EVs of all types seem as inconvenient as possible.

I can't recall seeing anything on the Monroney (EPA window) sticker that makes "EVs of all types seem as inconvenient as possible". What specifically are you referring to?

Actually, using MPGe results in numbers that look a lot higher than gasoline mpg numbers which sends the right overall message to less-informed consumers. For the rest of us, the sticker also includes kWh per 100 miles for electric and I think it may also include gallons per mile for gasoline.

 

[Red Sage]

I'm not following what you mean. They do not in any sense use "financially based" sets of numbers. They measure the kWh that go into the car from a wall socket. That is an energy consumption number. It has nothing directly to do with pollution emissions or energy cost. Those are calculated separately based on actual electricity sources.

When you go to the EPA's website and use the Compare Side-by-Side form, four sections at the bottom half of the form are financially related. Annual Fuel Cost, You Save or Spend, Cost to Drive 25 Miles, and Cost to Fill the Tank. At the bottom of the chart, you can Customize the results to your own costs using the Personalize link. Once there, by changing it to reflect your own relative costs for different grades of gasoline in your area, what you pay for electricity in your home, and the number and type of miles you drive per year, the calculated results will change based upon your own numbers when you click the Personalize button. The more expensive Premium gasoline is, the more it costs to operate a car per year. The more highway miles you drive, the better the overall fuel economy for most gas guzzlers.

I don't like MPGe either. I agree it is cryptic. They claim it was the best understood comparison metric among test groups of normals. Abnormal people like us would prefer Wh per mile.

Yeah! Wait... Abby 'something'... Abby Who? Abby... NORMAL? Hey! I resemble that remark!

I can't recall seeing anything on the Monroney (EPA window) sticker that makes "EVs of all types seem as inconvenient as possible". What specifically are you referring to?

Again, this is relative... If all the upstream costs of petroleum delivery were included in the equation, the advantage to EVs would be much higher. Either EVs would have an MPGe in the multiple hundreds... Or ICE vehicles would have 1/8th the MPG. It isn't fair that EVs are only allowed to be a mere twice as efficient as ICE cars. It is the maximum level of relatively viable inconvenience for them to be portrayed as such.

Actually, using MPGe results in numbers that look a lot higher than gasoline mpg numbers which sends the right overall message to less-informed consumers. For the rest of us, the sticker also includes kWh per 100 miles for electric and I think it may also include gallons per mile for gasoline.

Overall, it may work... But is still rather inaccurate. I would think a vehicle that holds the relative energy equivalent of 2-1/2 gallons of gasoline fuel, and achieves a 265 mile range with it, should be rated at 106 MPGe instead of 'only' 89 MPGe. That is a 16% efficiency penalty. Meanwhile, a vehicle with only 71% the energy capacity of a single gallon of gasoline and a range of 76 miles is awarded a 105 MPGe (Ford Focus Electric). And a car with a whopping huge 20 mile fully electric range is gifted an 88 MPGe (Ford C-MAX Energi). That is complete [BORSHT].

 

[SageBrush]

Again, this is relative... If all the upstream costs of petroleum delivery were included in the equation, the advantage to EVs would be much higher.

Compared to clean energy sourced EV, I agree
Compared to "average grid emissions," about 40 MPGe
Compared to "marginal emisssions," about 30 MPGe

 

[stopcrazypp]

From the Horse's mouth ...

A preview is available, or you can buy the full 26 page report for $74 ;-)

Addendum ... details. As usual, Jeff N was closest ...

RedSage was right -- the range estimate includes charging losses.

This is incorrect. The Range estimate does not include charging losses. All it simulates is running the given cycle until the battery can't continue (AKA usable DC kWh). If you know the usable kWh of the battery, the range figure gives you consumption not including charging losses.

The MPGe however, does include charging losses (if you use the range figure with the MPGe figure it gives you the AC kWh consumed at the EVSE).

Also, the actual updated test procedure (your link is from 2012) allows using shortcuts because the original full procedure takes too long for long range vehicles.
Stop the Press! Tesla announces REAL HP numbers for P85D and P90L

 

[SageBrush]

This is incorrect. The Range estimate does not include charging losses. All it simulates is running the given cycle until the battery can't continue (AKA usable DC kWh). If you know the usable kWh of the battery, the range figure gives you consumption not including charging losses.

Read the details link. It is government source. Below is the relevant snippet with my bolding

the battery is recharged from a normal AC source and the energy consumption of the vehicle is determined (in kW-hr/mile or kW-hr/100 miles) by dividing the kilowatt-hours of energy to recharge the battery by the miles traveled by the vehicle.

I thought exactly as you do, so consider both of us corrected.

 

[ItsNotAboutTheMoney]

Read the details link. It is government source. Below is the relevant snippet with my bolding

I thought exactly as you do, so consider both of us corrected.

The city driving range is determined from the number of miles driven over the city cycle until the vehicle can no longer follow the driving cycle.

...

The same test SAE J1634 procedure outlined above, is used to determine the highway energy consumption and the highway driving range (except the vehicle is operated over successive highway cycles).

...

• by multiplying city/highway fuel economy and range values by 0.7 and dividing city/highway energy consumption and CO2 values by 0.7;
• using the vehicle specific 5-cycle method described in 40 CFR 600.210-12(a)(1);
• using a method which is equivalent to the vehicle specific 5-cycle method described in 40 CFR
600.210-12(a)(1) (with prior EPA approval);
• using adjustment factors which are based on in-use data (with prior EPA approval).

Range is simply based on driving until the battery is exhausted. The range _may_ be _adjusted_, but it has _nothing_ to do with the method of charging, unless somehow AC charging is able to add more charge than DC charging.

Efficiency ratings are wall-to-wheel, range ranges are battery-to-wheel.

 

[stopcrazypp]

Read the details link. It is government source. Below is the relevant snippet with my bolding

I thought exactly as you do, so consider both of us corrected.

In the long form test you quoted, the AC consumption that you bolded is used to calculate the MPGe figure. It has nothing to do with the range figure (which is based solely on the cycling), even though both are captured in the same test procedure.

Like ItsNotAboutTheMoney says below, the energy consumed charging and type of charging is irrelevant to the range unless AC charging adds a different amount of charge (you end up with different usable DC kWh).

 

[SageBrush]

In the long form test you quoted, the AC consumption that you bolded is used to calculate the MPGe figure. It has nothing to do with the range figure (which is based solely on the cycling), even though both are captured in the same test procedure.

I read the procedure again and I think you are right. My bolding

Electric Vehicle - City Test Procedure Summary - Following SAE J1634 Recommended Practice, the battery is fully charged, the vehicle is parked over night, and then the following day the vehicle driven over successive city cycles until the battery becomes discharged (and the vehicle can no longer follow the city driving cycle). After running the successive city cycles, the battery is recharged from a normal AC source and the energy consumption of the vehicle is determined (in kW-hr/mile or kW-hr/100 miles) by dividing the kilowatt-hours of energy to recharge the battery by the miles traveled by the vehicle. To calculate the energy consumption in units of mpge (miles/gallon equivalent) we use a conversion factor of 33.705 kilowatt-hours of electricity per gallon of gasoline (which is basically a measure of the energy in gasoline (in BTUs) converted to electricity). The city driving range is determined from the number of miles driven over the city cycle until the vehicle can no longer follow the driving cycle.

The highway cycle follow the same method, then for range the average is multiplied by 0.7. This seems like a reasonable test method, although the 0.7 fudge factor leaves a lot of ymmv for individual drivers. It also ignores the practical issue of range mostly being a concern for long highway drives. Range for a car like the GM Bolt will likely over-estimate, and under-estimate the Model ☰ due to the included weighting of the city drive cycle.

There is a shortcut version that lets the testers discharge the battery only once rather than for each test but the approach is the same.

 

[omgwtfbyobbq]

I presume that there is a 10% 'buffer' for anti-bricking, and that EPA testing results also offer an additional 10% penalty. So, 75,000 Wh times 0.90 comes to 67,500 Wh available for use. 67,500 Wh times 0.90 is 60,750 Wh remaining, once you assume as the EPA does that you have 'spilled electrons on the floor' while filling your car in the garage. 60,500 Wh divided by 200 Wh per mile comes to 303.75 miles, which could be rounded up to 304 miles range.

As usual, I could be wrong. The percentage of the 'reserve' in the battery pack may be a bit less. The penalty for either drivetrain losses or induction losses during charging may be more. But I think this comes pretty close to what we have observed so far. In other words, I no longer count the entire expected capacity of a battery pack as being available for use.



By the way, this is also why I argue so vehemently against a 50 kWh, 45 kWh, or 40 kWh version of the Model ☰. Yeah, sure... If you could use every iota of that capacity, and tuned the vehicle to be incredibly efficient, they might achieve a range over 200 miles... 50 kWh would require 250 Wh or less consumption. 45 kWh at under 225 Wh per mile. 40 kWh at below 200 Wh per mile. Maybe. If you tuned the car to have the equivalent Performance profile of a Prius? It would not be too easy though, given what I have witnessed before.

Knowing that Tesla Motors originally expected ranges of 300-to-320 miles (Model S 85), 230 miles (Model S 60), and 160 miles (Model S 40), I'd rather not cut it so close. Because with the EPA's 5-Cycle Range Tests, the results were 265 miles (88.3% of expected range), 208 miles (90.4%), and 139 miles (86.8%) instead. If the EPA just happens to introduce a new range test, just ahead of the Model ☰ release, that is weighted in favor of ICE, Tesla Motors may not officially reach their intended minimum range. That would be bad.

I don't think range generalizes like that, but I could be wrong too. On the plus side, all these discussions encouraged me to throw together a quick python script that uses the actual EPA FTP-75/HWFET data and S 70D range to come up with a guesstimate for the range of a 55kWh and 75kWh pack. Feel free to provide constructive criticism and tweak the parameters of the model as you see fit.

Model 3 Specification Speculation

 

[wdolson]

I'm not sure about the EPA methodology but it makes sense to include the charging losses when the number is 'electricity used per mile,' but not when calculating range. It would also be interesting to know whether EPA uses the battery nominal capacity or the usable SOC.

I'll have to see whether EPA has numbers to reflect each situation. Or maybe the EPA always includes charging losses and does not account for unused SOC. If both are ~ 10% then the final number is fairly accurate, at least for Li-x.

Addendum:
I checked the 90D Mx. From fueleconomy.gov,
MPGe city 90 Highway 94, average 92
Consumption: 37 kWh/100 miles
92 MPGe average = 2.73 miles per kWh
= 36.3 kWh

So within a rounding error*, EPA uses the same assumptions for each number. The sticker says 257 miles range. If based on 37 kWh/100 miles, it implies 95 kWh usable.

My GUESS, then for range:
highway consumption
Charging losses
Nominal capacity rating

*
94.5 MPGe highway
94.5 miles / 33.7 kWh
= 2.8 miles / kWh
= 91.7 kWh for EPA range of 257 miles

It probably would be better to use the numbers for the Model S which are 33 KWh/100mi for the 90D and 35 KWh/100 mi for the P90D. The combined MPGe for the 90D is 103.

 

[Fred Thompson]

What Tesla is doing right now is an anomaly. They don't want to give away the extra 5 KWh on the 75 KWh packs, but they probably ran out of batteries for the 70 KWh packs before they filled all the 70 KWh orders, so they are doing this as a transition. They earlier limited the 60 pack to 40 because there were not enough orders for the 40 to justify engineering an all new pack. When the initial orders for the 40 pack were filled, they dropped it from the line up.

To make just a 100 KWh pack and then software limit it down to 75 or 70 would be very poor economics. The 75 KWh pack probably has the same number of cells as the 70, just higher density. The 85/90 have two extra modules the 70 doesn't have and building the packs with the extra batteries with no guarantee they will ever make any money off the extra capacity is foolish economics.

They can put some features in there and software limit it because the cost of installing the unused tech on a relative handful of cars probably doesn't cost as much as it would to have to maintain stock of two sets of parts. This is the case with the new 48/72 A charger. It's cheaper to have one part that can be turned on later than it is to maintain stock of two parts.

They probably figured their losses from the relative few cars that pass on AP is less than the trouble stocking parts for no AP cars. Even at that they are trying to entice people who passed on AP to buy it by giving them a free trial.

Batteries are expensive and the supply is limited. Except for times when Tesla is transitioning from one battery type to another, it doesn't make sense to build the largest battery and software limit it. I expect the 70 KWh will be dropped entirely in a few months.



The decreases in costs and increases in capacity average 5-10% a year, but that's only if you look long term. Year to year it's more fits and starts. People got used to Moore's Law which was a fairly consistent progression of capability and they expect that with batteries, but there is nothing anywhere close to Moore's Law for batteries. JB Straubel has said it many times in talks.

If you're doing long term planning, you can guess in 10 years capacity will be around X and price around Y, but predicting when the next break is going to be is more problematic.

While Tesla remains supply constrained, this is incorrect. Until that is addressed, nothing that they change will grow the number of purchases.

While Tesla remains supply constrained, this is incorrect. Until that is addressed, nothing that they change will grow the number of purchases.

Yes, Tesla would have to convince customers that they will be able to meet current demand before they can attract additional customers with an option to increase range with a software update. They would also have to crunch the numbers and determine that it is financially to their advantage to reduce the number of battery sizes by providing a battery that could be updated with software to increase vehicle range. Tesla is currently attempting to increase reservation by advertising people should order now if they want a Model 3 delivered in 2018.

 

[JRP3]

And a 90 kWh battery pack would occupy only 74% as much volume as an 85 kWh battery pack from 2012.

Of course in reality a 90kWh pack currently takes the same amount of space as the 85kWh pack. However I do expect to see a 100kWh pack in the near future, for the S/X.

 

[brianman]

Seems like extortion otherwise.

Seriously?

 

[brianman]

Yes, Tesla would have to convince customers that they will be able to meet current demand before they can attract additional customers ....

No, it's simpler than that. Until Tesla production saturates the demand, there can be no additional customers/sales no matter what they do.