Saturday, September 26, 2015

EV ranking

As with all engineering projects, EV is a matter of trade off. Any old fool can make an EV that gets 1000 miles range per charge or 0-60 mph time of 2 seconds with enough money. EV that cost millions of dollars each is not a good EV. You can't just look at single metric and say that's good (or bad) EV.

SparkEV does well in individual rating, such as second most efficient EV behind BMW i3, and third quickest EV from 0-60 mph behind Tesla S and i3. It is also lowest cost EV that gets 80 miles or more range. But how good is it with respect to cost? What we need is an objective metric to determine what good EV means with pricing as a factor. How objective are these? It's my blog, I can make up whatever I want. But I'll try to be fair.

Range-cost metric

Like stock price P/E ratio, range per dollar is helpful. I use post subsidy cost. Since most EV buyers take full subsidy, I call this the real world cost. Indeed, Tesla falls to the bottom half when measured this way. There are few web sites that discuss pre-subsidy range/$ that show Tesla to be the top followed by SparkEV. Such is nonsense, of course. Why wouldn't you take the subsidy? Buying EV without subsidy makes no sense when they are currently available.

To clarify such rampant misinformation, here's a ranking of some popular EV based on range/$ metric post subsidy. Subsidy is assumed to be $10K ($7.5K fed, $2.5K CA, although it could be as high as $4K for low income CA and $2250 out of $3K in MD). Only cars with DC fast charging are on this list. I consider EV without DCFC as toys, not serious vehicles. Yellow (jersey as in bicycle race) highlights the best.



"Toys" not worth considering

Fiat 500e
Ford Focus Electric: 76 miles / $20K = 3.8
Leaf S without DCFC option
Mercedes B class
Smart ForTwo EV
SparkEV without DCFC option: 82 miles / $15K = 5.47
Toyota Rav4EV: 113 miles / $40K = 2.83
Golf cart: 10 miles / $5K = 2.00
Used Golf cart with new battery: 10 miles / $2K = 5.00
Really beat up old golf cart with new battery: 5 miles / $1K = 5.00

Few interesting observations can be made.

1. SparkEV is clearly on top of the pack with good margin. SparkEV is even better than low cost golf cart in range-cost metric. Golf courses should just let people drive SparkEV instead of golf carts. Maybe GM could add "golf-cart mode" for just such purpose to limit speed and acceleration, although marketing might be problematic given how "sporty" SparkEV is.

2. 2016 Leaf SV/SL with longer range is something to behold. It's the first EV under $30K to break the 100 miles per charge barrier. With widespread Chademo DCFC, it's probably second best EV on the market today (SparkEV being top). If you're not in moderate weather areas (eg. AZ), Kia Soul EV would be a better choice due to its thermal management.

3. Which brings to attention Kia Soul EV. I test drove it in Drive Electric event, and it is ok. Being used to mashing the accelerator in SparkEV and feeling the rush, Soul EV lacked soul. If it's not for Leaf SV/SL 110 miles range, Kia Soul EV would take the second place.

4. Tesla S70 is a surprise. I had expected it to be in top half, but it's in bottom half. In that regard, it doesn't make sense to get base model S. Either go for P90D for performance or skip Tesla altogether.

5. BMW i3 is a huge surprise to me. It's a great car with good performance and features similar to SparkEV. Alas, its high price makes it, well, high priced. Long ago, I had considered BMW i3 to be better version of SparkEV, but it seems that's not the case; SparkEV is leagues ahead of BMW i3.

6. At the bottom, I add Bolt / Model 3, both of which are vaporware for now. I mean, Chevy won't even let you sit in one at autoshow. Still, this shows how SparkEV would hold up in the future: not as good, but that's expected.

Clearly Bolt/Model 3 wins over SparkEV, right? But hold on. While exact spec isn't known, one can guess that battery will be bigger (50kWh?), hence the replacement cost to be at least $5,000 after warranty expires after 10 years. This assumes battery prices have come down to $100/kWh (from about $300/kWh today) after 10 years, probably the best case scenario considering that raw material prices are around $100/kWh. One can look today to see how many people spend $5,000 to fix a broken down 10 year old car: not many, if at all. In comparison, SparkEV with 19.5 kWh battery could cost less than $2,000. Lots of people spend $2000 to fix 10 year old car.

What this comes down to is the resale value of 10 year old car. In case of Bolt/Model 3, car is basically junk. Entire cost is sunk. But in case of SparkEV, car would be fixable with reasonable cost and re-sold as used and driven for another 10 years. One can play the game that used battery could cost less or that shorter range could still be sold as used car. Well, I guess we'll have to wait and see: definite lower practical cost of SparkEV or some nebulous hocus pocus for much bigger battery EV. In my opinion, bird-in-hand wins even over vaporware promise, so SparkEV still wins the crown as best EV, even against 200 miles range per charge EV.

Performance-cost metric

One can also try another cost metric, such as $ to 0-60 mph time. Again, one can put 9999 horsepower motor with 1 kWh battery and get 0-60 mph in 2.2 seconds, but that probably won't cost $16K. As expected, SparkEV wins this one hands down, too. At this point in time, only SparkEV and Tesla P90D are worth discussing as they are only two that can compete (and beat) other cars in their price range, but let's see few others. BMW comes close, but there are other cars in its price range that are quicker. I also add Corvette in the mix, a decent gas car with limited utility. Metric is time-$ product where smaller is better. Yellow (jersey as in bicycle race) highlights the best. Let the battle begin!



As I was putting together the data, I noticed SparkEV is far above anything else as shown in 0-60 time-cost product (almost 2.6 times more bang for the buck than P90D and Corvette!). I decided to square the time to give advantage to P90D. Hey, it's my blog, I'll tweak the figure to favor another vehicle if I want to! But even then, SparkEV comes out ahead. Only when you cube the time does P90D gets better than SparkEV.

Range-performance-cost

Let's try something little more convoluted by combining 3 metrics. EV with longer range, short 0-60 time, low cost is good. Then the equation used is 

R / (t * $) where R=range, t=0 to 60 time in seconds, $ is real cost post subsidy.

where resulting larger number is better EV.


This formula produces result that I like. SparkEV, like other cost metrics, comes out on top. Yawn; what else is new? But Tesla P90D is close second. (This was when I erroneously used $145K for P90D. Despite being second fiddle to P90D, it's nipping at its heels!)

Other popular EV are clustered around 400 mark. A surprise is VW eGolf worse than i-MiEV. Maybe it's not a huge surprise given that it's very similar to Leaf with higher cost.

Another interesting bit is Bolt/Model3. While specs aren't out, it could be roughly on par with P90D. (Again, when I wrongly had P90D at $135K) If they can bring the price down (not likely) or quicker 0-60, it could be competitive to SparkEV. How quick? 9.3 seconds. That should be doable. But even that won't compare to SparkEV single metric of 7.2 seconds. Would you buy $30K car that performs poorer than $16K car? I think I'm going to be keeping my SparkEV even after Bolt/Model3 comes out.

Range-performance-battery-cost

As discussed before, smaller battery is better due to potential for lower replacement cost. While it doesn't impact the car when new, it will have large effect when it comes time to replace it 10 years later. This also affects whether the car will end up in the junk yard (more pollution and energy use) or repaired and continue to drive.

R / (t * B * $) where R=range, t=0 to 60 time in seconds, B is battery capacity in kWh, $ is real cost post subsidy.


Yet again, SparkEV wins. Not surprisingly, Tesla P90D is at the bottom due to its large battery. The question for P90D owners is if they're willing to spend $9000+ for a 10 year old car? Of course, Tesla offers optional battery replacement plan, but I don't know how many take up on their offer and continue to pay for 10 years. Given that P90D is not a typical consumer car, I suspect many will opt to keep it running by spending $9000+, and not throw it in the junk yard.

In case of SparkEV, like P90D, it can out accelerate comparable cost gas cars, so spending $2000 to replace the battery may be worth it. I have a feeling SparkEV will become collector's car if it's not mass produced by Chevy.

But what about other EV that are inferior performing to comparable gas cars like SoulEV and Leaf SV/SL or even i3? Would you spend $3000 for a 10 year old car that's slower than gas car counterparts and doesn't even have liquid cooled thermal management in case of Leaf, especially when there will be new EV with 200+ miles range that perform better? I have a feeling they will end up in junk yard, causing more pollution and energy waste than comparable gas car where you can spend couple of thousand dollars incrementally to drive another 10 years rather than one lump sum battery replacement.

Other-cost metrics to consider

Another comparison could be efficiency-cost metric, and one can see SparkEV wins this hands down due to being second most efficient EV (119 MPGe) behind BMW i3 (124 MPGe) while costing less than half.

Unfortunately, I don't have ready source of information on other metrics such as skid pad and slalom data. Considering SparkEV battery weighs less than 500lb and sitting low (typical EV), I suspect it would handle well, too, although stock tires would have to be changed.

Still, I'm curious what legitimate cost metric SparkEV would be less than another EV. Taking the cube of 0-60 mph time doesn't count. 3.3kW charger vs 6.6kW charger would be one, but that means little when DCFC is available. 4 hours for 24kWh Leaf at 6.6kW vs 6 hours for 19kWh SparkEV at 3.3kW makes virtually no difference when they are used at work (8 hours) or home (12 hours). In fact, it would be better if there's 2.2kW charger for SparkEV so that I can leave the car plugged in while at work rather than moving it half way through the work day.

Actually, there is one metric where SparkEV definitely lose. It's been sold out for many months! Maybe it's a sign that SparkEV is destined to become a collector's car.

Edit Oct. 28, 2015

Nissan Leaf (even 2016 with 110 miles range) does not have thermal mangement. That results in poor battery longevity, and very poor DC fast charge performance. I've seen several Leaf charging at 2kW out of 50kW fast chargers! Such slow charging is bad for Leaf drivers, but also bad for all EV who have to wait for Leaf to charge. I've harped on this many times in my blog.

http://sparkev.blogspot.com/2015/05/vs-nissan-leaf-quick-charge.html

http://sparkev.blogspot.com/2015/10/free-charging-sucks.html

http://sparkev.blogspot.com/2015/10/jerks-all-around-us-iced-leafed.html

It has come to my attention that VW eGolf also  does not have thermal management. While I don't know if their DCFC would be as slow as Leaf since I've only seen one. But what I know is that lack of thermal management is a risk, not onlyb to battery longevity, but also wasted time at fast chargers both for the eGolf driver and everyone else. One should weigh this "problem" appropriately. If it's up to me, it doesn't make sense to get eGolf with similar specs as Leaf S for $3000 more, but then again, no other EV makes sense for me other than SparkEV.

Below are some links to eGolf's lack of thermal management.

http://www.autoblog.com/2014/03/31/vw-e-golf-will-not-have-active-cooling-system-lithium-battery/

http://www.greencarcongress.com/2014/07/20140721-egolf.html

Edit Jan. 31, 2016

While Rav4EV was considered a "toy" above due to lack of DCFC, someone thought it doesn't have to be that way. Tony Williams of quickchargepower.com offers Chademo charging modification to Rav4EV for $3000.

http://www.quickchargepower.com

That would make Rav4EV pretty compelling. It has small SUV form, range of 113 miles using 42 kWh battery, 0-60 in 7.2 seconds (some say even quicker). While Toyota discontinued Rav4EV at 2014 model year, used Rav4EV are selling for about $25K to $30K. Assuming $30K for used price including Chademo mod, some scores can be given.

Range / price = 113 / 30 = 3.77 (roughly Tesla S70)

Performance * price = 7.2 * 30 = 216 (7.2*7.2*30 = 1555, roughly BMW i3)

Range / (Performance * price) = 113 / 216 * 1000 = 523 (worse than SparkEV, but significantly better than  110 mile range 2016 Leaf SV/SL)

Range / (Performance * price * battery) = 523 / 42 = 12.5 (roughly eGolf)

Unfortunately, there are many forum posts (including Tony's) that state Rav4EV suffers from multiple problems, and it may have reliability issues. Since Toyota seem to be abandoning battery EV in favor of other (fuel cell for now), few drivers with Rav4EV seem to be left out in the cold.

More unfortunately, quickchargepower.com website seem to be out of order. However, I have seen some recent plugshare posts showing success with Rav4EV modified to use Chademo. The user name was Tony, so maybe it was Tony Williams? In any case, used Rav4EV under knowledgeable driver could make it a decent EV with quickchargepower.com's Chademo mod.

Bolt news

Chevy announced Bolt as 60 kWh battery, 0-60 under 7 seconds, about $30K. One has to guess what "under 7 seconds" means, but SparkEV was promised under 8 seconds to be 7.5 seconds in 2014 model. Let's guess 6.5 seconds for Bolt. That would make the scores as follows.

Range / price = 200 / 30 = 6.66 (same as my guess)

Performance * price = 6.5 * 30 = 192 (6.5*6.5*27 = 1268, roughly BMW i3)

Range / (Performance * price) = 200 / 192 * 1000 = 1042 (the best among EV)

Range / (Performance * price * battery) = 1042 / 60 = 17.4 (roughly BMW i3)

Of course, these are just guesses. But something that I did not include in the metric is DCFC time. SparkEV is quickest charging EV in the world with 0-80% in 20 minutes. By contrast, Tesla takes nearly 45 minutes for 0-80%. While one might poo-poo percentage as meaningless, that isn't the case when human psychology is involved.

http://sparkev.blogspot.com/2015/12/sparkev-is-quickest-charging-ev-in-world.html

Chevy announced that they have no intention of helping expand the DCFC network. Then existing 50kW DCFC chargers are to be used with 60 kWh battery Bolt, which would need close to an hour for DCFC. This would make Bolt the slowest fast charging EV in the world. This is not good.

For people who are looking to buy EV and only EV (no gas car) regardless of lack of performance, Bolt will be very good. Bolt will surely take market share away from Leaf, BMW i3, eGolf that cost similar as Bolt with far less range.

But for general public (mass market), Bolt isn't very good. With $30K, there are far better options. For example, Subaru WRX is quicker to 60 MPH AND it comes with AWD. Bolt falls way short.

http://sparkev.blogspot.com/2016/01/mass-market-ev-to-bolt-or-not-to-bolt.html

Therefore, SparkEV is still the best, not only as EV, but as a car. We'll revisit this when Model 3 is released. In theory, Model 3 could have all the advantages of SparkEV (quicker than comparable cost gas cars, charge in 20 minutes or less) while also having 200 miles range.

http://sparkev.blogspot.com/2016/01/mass-market-ev-hoping-for-tesla.html

1 comment:

  1. You should look at 0 to 40 acceleration on city cars 0 to 60 is meaningless for these cars

    ReplyDelete